fix(routing): pass api_mode from cheap_model config in smart routing by BukeLy · Pull Request #8321 · NousResearch/hermes-agent

BukeLy · 2026-04-12T11:06:21Z

What does this PR do?

resolve_runtime_provider does not accept an api_mode parameter. When smart model routing routes a message to the cheap_model, the configured api_mode (e.g. anthropic_messages) is silently dropped, and the runtime defaults to chat_completions.

This breaks local inference servers that require a specific API format, particularly when using Anthropic-format endpoints with SSD KV caching where api_mode determines cache hit behavior.

Fix: prefer route.get("api_mode") over runtime.get("api_mode") in the cheap model return path.

Related Issue

Fixes #8515

Type of Change

Bug fix (non-breaking change that fixes an issue)

Changes Made

agent/smart_model_routing.py: Added effective_api_mode that prefers route config over runtime, used in both runtime dict and signature tuple
tests/agent/test_smart_model_routing.py: Added 3 tests for api_mode priority, fallback, and signature inclusion

How to Test

Configure cheap_model.api_mode: anthropic_messages in smart routing config
Send a simple message that routes to cheap model
Verify the API call uses anthropic_messages format (not chat_completions)

Checklist

Code

I've read the Contributing Guide
My commit messages follow Conventional Commits
I searched for existing PRs to make sure this isn't a duplicate
My PR contains only changes related to this fix
I've run pytest tests/ -q and all tests pass
I've added tests for my changes
I've tested on my platform: macOS 15.4 (Sequoia), Apple Silicon

Documentation & Housekeeping

N/A - no config keys added, existing cheap_model.api_mode now works as expected
Cross-platform: all platforms affected equally, fix is platform-independent

resolve_runtime_provider does not accept an api_mode parameter, so the cheap_model's configured api_mode (e.g. anthropic_messages for local servers with SSD KV cache) was silently dropped. The runtime always fell back to the default chat_completions mode. Prefer the route config's api_mode over the runtime-resolved value so custom endpoints that require a specific API format work correctly when reached via smart model routing.

Verify that api_mode from cheap_model config takes priority over runtime-resolved value, falls back correctly when unset, and is reflected in the routing signature for cache keying.

BukeLy · 2026-04-13T03:16:13Z

Hi @jbarket @chalkers @leon @intertwine,
could you please take a look at this PR and help enable the CI run? Thanks!

@nocoo

…pport Third-party gateways that speak the native Anthropic protocol (MiniMax, Zhipu GLM, Alibaba DashScope, Kimi, LiteLLM proxies) now work end-to-end with the same feature set as direct api.anthropic.com callers. Synthesizes eight stale community PRs into one consolidated change. Five fixes: - URL detection: consolidate three inline `endswith("/anthropic")` checks in runtime_provider.py into the shared _detect_api_mode_for_url helper. Third-party /anthropic endpoints now auto-resolve to api_mode=anthropic_messages via one code path instead of three. - OAuth leak-guard: all five sites that assign `_is_anthropic_oauth` (__init__, switch_model, _try_refresh_anthropic_client_credentials, _swap_credential, _try_activate_fallback) now gate on `provider == "anthropic"` so a stale ANTHROPIC_TOKEN never trips Claude-Code identity injection on third-party endpoints. Previously only 2 of 5 sites were guarded. - Prompt caching: new method `_anthropic_prompt_cache_policy()` returns `(should_cache, use_native_layout)` per endpoint. Replaces three inline conditions and the `native_anthropic=(api_mode=='anthropic_messages')` call-site flag. Native Anthropic and third-party Anthropic gateways both get the native cache_control layout; OpenRouter gets envelope layout. Layout is persisted in `_primary_runtime` so fallback restoration preserves the per-endpoint choice. - Auxiliary client: `_try_custom_endpoint` honors `api_mode=anthropic_messages` and builds `AnthropicAuxiliaryClient` instead of silently downgrading to an OpenAI-wire client. Degrades gracefully to OpenAI-wire when the anthropic SDK isn't installed. - Config hygiene: `_update_config_for_provider` (hermes_cli/auth.py) clears stale `api_key`/`api_mode` when switching to a built-in provider, so a previous MiniMax custom endpoint's credentials can't leak into a later OpenRouter session. - Truncation continuation: length-continuation and tool-call-truncation retry now cover `anthropic_messages` in addition to `chat_completions` and `bedrock_converse`. Reuses the existing `_build_assistant_message` path via `normalize_anthropic_response()` so the interim message shape is byte-identical to the non-truncated path. Tests: 6 new files, 42 test cases. Targeted run + tests/run_agent, tests/agent, tests/hermes_cli all pass (4554 passed). Synthesized from (credits preserved via Co-authored-by trailers): #7410 @nocoo — URL detection helper #7393 @keyuyuan — OAuth 5-site guard #7367 @n-WN — OAuth guard (narrower cousin, kept comment) #8636 @sgaofen — caching helper + native-vs-proxy layout split #10954 @Only-Code-A — caching on anthropic_messages+Claude #7648 @zhongyueming1121 — aux client anthropic_messages branch #6096 @hansnow — /model switch clears stale api_mode #9691 @TroyMitchell911 — anthropic_messages truncation continuation Closes: #7366, #8294 (third-party Anthropic identity + caching). Supersedes: #7410, #7367, #7393, #8636, #10954, #7648, #6096, #9691. Rejects: #9621 (OpenAI-wire caching with incomplete blocklist — risky), #7242 (superseded by #9691, stale branch), #8321 (targets smart_model_routing which was removed in #12732). Co-authored-by: nocoo <nocoo@users.noreply.github.com> Co-authored-by: Keyu Yuan <leoyuan0099@gmail.com> Co-authored-by: Zoee <30841158+n-WN@users.noreply.github.com> Co-authored-by: sgaofen <135070653+sgaofen@users.noreply.github.com> Co-authored-by: Only-Code-A <bxzt2006@163.com> Co-authored-by: zhongyueming <mygamez@163.com> Co-authored-by: Xiaohan Li <hansnow@users.noreply.github.com> Co-authored-by: Troy Mitchell <i@troy-y.org>

@nocoo

…pport (#12846) Third-party gateways that speak the native Anthropic protocol (MiniMax, Zhipu GLM, Alibaba DashScope, Kimi, LiteLLM proxies) now work end-to-end with the same feature set as direct api.anthropic.com callers. Synthesizes eight stale community PRs into one consolidated change. Five fixes: - URL detection: consolidate three inline `endswith("/anthropic")` checks in runtime_provider.py into the shared _detect_api_mode_for_url helper. Third-party /anthropic endpoints now auto-resolve to api_mode=anthropic_messages via one code path instead of three. - OAuth leak-guard: all five sites that assign `_is_anthropic_oauth` (__init__, switch_model, _try_refresh_anthropic_client_credentials, _swap_credential, _try_activate_fallback) now gate on `provider == "anthropic"` so a stale ANTHROPIC_TOKEN never trips Claude-Code identity injection on third-party endpoints. Previously only 2 of 5 sites were guarded. - Prompt caching: new method `_anthropic_prompt_cache_policy()` returns `(should_cache, use_native_layout)` per endpoint. Replaces three inline conditions and the `native_anthropic=(api_mode=='anthropic_messages')` call-site flag. Native Anthropic and third-party Anthropic gateways both get the native cache_control layout; OpenRouter gets envelope layout. Layout is persisted in `_primary_runtime` so fallback restoration preserves the per-endpoint choice. - Auxiliary client: `_try_custom_endpoint` honors `api_mode=anthropic_messages` and builds `AnthropicAuxiliaryClient` instead of silently downgrading to an OpenAI-wire client. Degrades gracefully to OpenAI-wire when the anthropic SDK isn't installed. - Config hygiene: `_update_config_for_provider` (hermes_cli/auth.py) clears stale `api_key`/`api_mode` when switching to a built-in provider, so a previous MiniMax custom endpoint's credentials can't leak into a later OpenRouter session. - Truncation continuation: length-continuation and tool-call-truncation retry now cover `anthropic_messages` in addition to `chat_completions` and `bedrock_converse`. Reuses the existing `_build_assistant_message` path via `normalize_anthropic_response()` so the interim message shape is byte-identical to the non-truncated path. Tests: 6 new files, 42 test cases. Targeted run + tests/run_agent, tests/agent, tests/hermes_cli all pass (4554 passed). Synthesized from (credits preserved via Co-authored-by trailers): #7410 @nocoo — URL detection helper #7393 @keyuyuan — OAuth 5-site guard #7367 @n-WN — OAuth guard (narrower cousin, kept comment) #8636 @sgaofen — caching helper + native-vs-proxy layout split #10954 @Only-Code-A — caching on anthropic_messages+Claude #7648 @zhongyueming1121 — aux client anthropic_messages branch #6096 @hansnow — /model switch clears stale api_mode #9691 @TroyMitchell911 — anthropic_messages truncation continuation Closes: #7366, #8294 (third-party Anthropic identity + caching). Supersedes: #7410, #7367, #7393, #8636, #10954, #7648, #6096, #9691. Rejects: #9621 (OpenAI-wire caching with incomplete blocklist — risky), #7242 (superseded by #9691, stale branch), #8321 (targets smart_model_routing which was removed in #12732). Co-authored-by: nocoo <nocoo@users.noreply.github.com> Co-authored-by: Keyu Yuan <leoyuan0099@gmail.com> Co-authored-by: Zoee <30841158+n-WN@users.noreply.github.com> Co-authored-by: sgaofen <135070653+sgaofen@users.noreply.github.com> Co-authored-by: Only-Code-A <bxzt2006@163.com> Co-authored-by: zhongyueming <mygamez@163.com> Co-authored-by: Xiaohan Li <hansnow@users.noreply.github.com> Co-authored-by: Troy Mitchell <i@troy-y.org>

teknium1 · 2026-04-20T05:44:03Z

Unfortunately agent/smart_model_routing.py was removed from the codebase in PR #12732 (commit 424e9f3) — the whole smart_model_routing feature was refactored out. This PR's fix targets code that no longer exists. Thanks for the contribution regardless; the api_mode persistence issue it raised was covered in a different layer by PR #12846 (config-time _update_config_for_provider + runtime _detect_api_mode_for_url helper).

Fix shipped in: #12846

@nocoo

…pport (NousResearch#12846) Third-party gateways that speak the native Anthropic protocol (MiniMax, Zhipu GLM, Alibaba DashScope, Kimi, LiteLLM proxies) now work end-to-end with the same feature set as direct api.anthropic.com callers. Synthesizes eight stale community PRs into one consolidated change. Five fixes: - URL detection: consolidate three inline `endswith("/anthropic")` checks in runtime_provider.py into the shared _detect_api_mode_for_url helper. Third-party /anthropic endpoints now auto-resolve to api_mode=anthropic_messages via one code path instead of three. - OAuth leak-guard: all five sites that assign `_is_anthropic_oauth` (__init__, switch_model, _try_refresh_anthropic_client_credentials, _swap_credential, _try_activate_fallback) now gate on `provider == "anthropic"` so a stale ANTHROPIC_TOKEN never trips Claude-Code identity injection on third-party endpoints. Previously only 2 of 5 sites were guarded. - Prompt caching: new method `_anthropic_prompt_cache_policy()` returns `(should_cache, use_native_layout)` per endpoint. Replaces three inline conditions and the `native_anthropic=(api_mode=='anthropic_messages')` call-site flag. Native Anthropic and third-party Anthropic gateways both get the native cache_control layout; OpenRouter gets envelope layout. Layout is persisted in `_primary_runtime` so fallback restoration preserves the per-endpoint choice. - Auxiliary client: `_try_custom_endpoint` honors `api_mode=anthropic_messages` and builds `AnthropicAuxiliaryClient` instead of silently downgrading to an OpenAI-wire client. Degrades gracefully to OpenAI-wire when the anthropic SDK isn't installed. - Config hygiene: `_update_config_for_provider` (hermes_cli/auth.py) clears stale `api_key`/`api_mode` when switching to a built-in provider, so a previous MiniMax custom endpoint's credentials can't leak into a later OpenRouter session. - Truncation continuation: length-continuation and tool-call-truncation retry now cover `anthropic_messages` in addition to `chat_completions` and `bedrock_converse`. Reuses the existing `_build_assistant_message` path via `normalize_anthropic_response()` so the interim message shape is byte-identical to the non-truncated path. Tests: 6 new files, 42 test cases. Targeted run + tests/run_agent, tests/agent, tests/hermes_cli all pass (4554 passed). Synthesized from (credits preserved via Co-authored-by trailers): NousResearch#7410 @nocoo — URL detection helper NousResearch#7393 @keyuyuan — OAuth 5-site guard NousResearch#7367 @n-WN — OAuth guard (narrower cousin, kept comment) NousResearch#8636 @sgaofen — caching helper + native-vs-proxy layout split NousResearch#10954 @Only-Code-A — caching on anthropic_messages+Claude NousResearch#7648 @zhongyueming1121 — aux client anthropic_messages branch NousResearch#6096 @hansnow — /model switch clears stale api_mode NousResearch#9691 @TroyMitchell911 — anthropic_messages truncation continuation Closes: NousResearch#7366, NousResearch#8294 (third-party Anthropic identity + caching). Supersedes: NousResearch#7410, NousResearch#7367, NousResearch#7393, NousResearch#8636, NousResearch#10954, NousResearch#7648, NousResearch#6096, NousResearch#9691. Rejects: NousResearch#9621 (OpenAI-wire caching with incomplete blocklist — risky), NousResearch#7242 (superseded by NousResearch#9691, stale branch), NousResearch#8321 (targets smart_model_routing which was removed in NousResearch#12732). Co-authored-by: nocoo <nocoo@users.noreply.github.com> Co-authored-by: Keyu Yuan <leoyuan0099@gmail.com> Co-authored-by: Zoee <30841158+n-WN@users.noreply.github.com> Co-authored-by: sgaofen <135070653+sgaofen@users.noreply.github.com> Co-authored-by: Only-Code-A <bxzt2006@163.com> Co-authored-by: zhongyueming <mygamez@163.com> Co-authored-by: Xiaohan Li <hansnow@users.noreply.github.com> Co-authored-by: Troy Mitchell <i@troy-y.org>

@nocoo

…pport (NousResearch#12846) Third-party gateways that speak the native Anthropic protocol (MiniMax, Zhipu GLM, Alibaba DashScope, Kimi, LiteLLM proxies) now work end-to-end with the same feature set as direct api.anthropic.com callers. Synthesizes eight stale community PRs into one consolidated change. Five fixes: - URL detection: consolidate three inline `endswith("/anthropic")` checks in runtime_provider.py into the shared _detect_api_mode_for_url helper. Third-party /anthropic endpoints now auto-resolve to api_mode=anthropic_messages via one code path instead of three. - OAuth leak-guard: all five sites that assign `_is_anthropic_oauth` (__init__, switch_model, _try_refresh_anthropic_client_credentials, _swap_credential, _try_activate_fallback) now gate on `provider == "anthropic"` so a stale ANTHROPIC_TOKEN never trips Claude-Code identity injection on third-party endpoints. Previously only 2 of 5 sites were guarded. - Prompt caching: new method `_anthropic_prompt_cache_policy()` returns `(should_cache, use_native_layout)` per endpoint. Replaces three inline conditions and the `native_anthropic=(api_mode=='anthropic_messages')` call-site flag. Native Anthropic and third-party Anthropic gateways both get the native cache_control layout; OpenRouter gets envelope layout. Layout is persisted in `_primary_runtime` so fallback restoration preserves the per-endpoint choice. - Auxiliary client: `_try_custom_endpoint` honors `api_mode=anthropic_messages` and builds `AnthropicAuxiliaryClient` instead of silently downgrading to an OpenAI-wire client. Degrades gracefully to OpenAI-wire when the anthropic SDK isn't installed. - Config hygiene: `_update_config_for_provider` (hermes_cli/auth.py) clears stale `api_key`/`api_mode` when switching to a built-in provider, so a previous MiniMax custom endpoint's credentials can't leak into a later OpenRouter session. - Truncation continuation: length-continuation and tool-call-truncation retry now cover `anthropic_messages` in addition to `chat_completions` and `bedrock_converse`. Reuses the existing `_build_assistant_message` path via `normalize_anthropic_response()` so the interim message shape is byte-identical to the non-truncated path. Tests: 6 new files, 42 test cases. Targeted run + tests/run_agent, tests/agent, tests/hermes_cli all pass (4554 passed). Synthesized from (credits preserved via Co-authored-by trailers): NousResearch#7410 @nocoo — URL detection helper NousResearch#7393 @keyuyuan — OAuth 5-site guard NousResearch#7367 @n-WN — OAuth guard (narrower cousin, kept comment) NousResearch#8636 @sgaofen — caching helper + native-vs-proxy layout split NousResearch#10954 @Only-Code-A — caching on anthropic_messages+Claude NousResearch#7648 @zhongyueming1121 — aux client anthropic_messages branch NousResearch#6096 @hansnow — /model switch clears stale api_mode NousResearch#9691 @TroyMitchell911 — anthropic_messages truncation continuation Closes: NousResearch#7366, NousResearch#8294 (third-party Anthropic identity + caching). Supersedes: NousResearch#7410, NousResearch#7367, NousResearch#7393, NousResearch#8636, NousResearch#10954, NousResearch#7648, NousResearch#6096, NousResearch#9691. Rejects: NousResearch#9621 (OpenAI-wire caching with incomplete blocklist — risky), NousResearch#7242 (superseded by NousResearch#9691, stale branch), NousResearch#8321 (targets smart_model_routing which was removed in NousResearch#12732). Co-authored-by: nocoo <nocoo@users.noreply.github.com> Co-authored-by: Keyu Yuan <leoyuan0099@gmail.com> Co-authored-by: Zoee <30841158+n-WN@users.noreply.github.com> Co-authored-by: sgaofen <135070653+sgaofen@users.noreply.github.com> Co-authored-by: Only-Code-A <bxzt2006@163.com> Co-authored-by: zhongyueming <mygamez@163.com> Co-authored-by: Xiaohan Li <hansnow@users.noreply.github.com> Co-authored-by: Troy Mitchell <i@troy-y.org>

@nocoo

…pport (NousResearch#12846) Third-party gateways that speak the native Anthropic protocol (MiniMax, Zhipu GLM, Alibaba DashScope, Kimi, LiteLLM proxies) now work end-to-end with the same feature set as direct api.anthropic.com callers. Synthesizes eight stale community PRs into one consolidated change. Five fixes: - URL detection: consolidate three inline `endswith("/anthropic")` checks in runtime_provider.py into the shared _detect_api_mode_for_url helper. Third-party /anthropic endpoints now auto-resolve to api_mode=anthropic_messages via one code path instead of three. - OAuth leak-guard: all five sites that assign `_is_anthropic_oauth` (__init__, switch_model, _try_refresh_anthropic_client_credentials, _swap_credential, _try_activate_fallback) now gate on `provider == "anthropic"` so a stale ANTHROPIC_TOKEN never trips Claude-Code identity injection on third-party endpoints. Previously only 2 of 5 sites were guarded. - Prompt caching: new method `_anthropic_prompt_cache_policy()` returns `(should_cache, use_native_layout)` per endpoint. Replaces three inline conditions and the `native_anthropic=(api_mode=='anthropic_messages')` call-site flag. Native Anthropic and third-party Anthropic gateways both get the native cache_control layout; OpenRouter gets envelope layout. Layout is persisted in `_primary_runtime` so fallback restoration preserves the per-endpoint choice. - Auxiliary client: `_try_custom_endpoint` honors `api_mode=anthropic_messages` and builds `AnthropicAuxiliaryClient` instead of silently downgrading to an OpenAI-wire client. Degrades gracefully to OpenAI-wire when the anthropic SDK isn't installed. - Config hygiene: `_update_config_for_provider` (hermes_cli/auth.py) clears stale `api_key`/`api_mode` when switching to a built-in provider, so a previous MiniMax custom endpoint's credentials can't leak into a later OpenRouter session. - Truncation continuation: length-continuation and tool-call-truncation retry now cover `anthropic_messages` in addition to `chat_completions` and `bedrock_converse`. Reuses the existing `_build_assistant_message` path via `normalize_anthropic_response()` so the interim message shape is byte-identical to the non-truncated path. Tests: 6 new files, 42 test cases. Targeted run + tests/run_agent, tests/agent, tests/hermes_cli all pass (4554 passed). Synthesized from (credits preserved via Co-authored-by trailers): NousResearch#7410 @nocoo — URL detection helper NousResearch#7393 @keyuyuan — OAuth 5-site guard NousResearch#7367 @n-WN — OAuth guard (narrower cousin, kept comment) NousResearch#8636 @sgaofen — caching helper + native-vs-proxy layout split NousResearch#10954 @Only-Code-A — caching on anthropic_messages+Claude NousResearch#7648 @zhongyueming1121 — aux client anthropic_messages branch NousResearch#6096 @hansnow — /model switch clears stale api_mode NousResearch#9691 @TroyMitchell911 — anthropic_messages truncation continuation Closes: NousResearch#7366, NousResearch#8294 (third-party Anthropic identity + caching). Supersedes: NousResearch#7410, NousResearch#7367, NousResearch#7393, NousResearch#8636, NousResearch#10954, NousResearch#7648, NousResearch#6096, NousResearch#9691. Rejects: NousResearch#9621 (OpenAI-wire caching with incomplete blocklist — risky), NousResearch#7242 (superseded by NousResearch#9691, stale branch), NousResearch#8321 (targets smart_model_routing which was removed in NousResearch#12732). Co-authored-by: nocoo <nocoo@users.noreply.github.com> Co-authored-by: Keyu Yuan <leoyuan0099@gmail.com> Co-authored-by: Zoee <30841158+n-WN@users.noreply.github.com> Co-authored-by: sgaofen <135070653+sgaofen@users.noreply.github.com> Co-authored-by: Only-Code-A <bxzt2006@163.com> Co-authored-by: zhongyueming <mygamez@163.com> Co-authored-by: Xiaohan Li <hansnow@users.noreply.github.com> Co-authored-by: Troy Mitchell <i@troy-y.org>

@nocoo

…pport (NousResearch#12846) Third-party gateways that speak the native Anthropic protocol (MiniMax, Zhipu GLM, Alibaba DashScope, Kimi, LiteLLM proxies) now work end-to-end with the same feature set as direct api.anthropic.com callers. Synthesizes eight stale community PRs into one consolidated change. Five fixes: - URL detection: consolidate three inline `endswith("/anthropic")` checks in runtime_provider.py into the shared _detect_api_mode_for_url helper. Third-party /anthropic endpoints now auto-resolve to api_mode=anthropic_messages via one code path instead of three. - OAuth leak-guard: all five sites that assign `_is_anthropic_oauth` (__init__, switch_model, _try_refresh_anthropic_client_credentials, _swap_credential, _try_activate_fallback) now gate on `provider == "anthropic"` so a stale ANTHROPIC_TOKEN never trips Claude-Code identity injection on third-party endpoints. Previously only 2 of 5 sites were guarded. - Prompt caching: new method `_anthropic_prompt_cache_policy()` returns `(should_cache, use_native_layout)` per endpoint. Replaces three inline conditions and the `native_anthropic=(api_mode=='anthropic_messages')` call-site flag. Native Anthropic and third-party Anthropic gateways both get the native cache_control layout; OpenRouter gets envelope layout. Layout is persisted in `_primary_runtime` so fallback restoration preserves the per-endpoint choice. - Auxiliary client: `_try_custom_endpoint` honors `api_mode=anthropic_messages` and builds `AnthropicAuxiliaryClient` instead of silently downgrading to an OpenAI-wire client. Degrades gracefully to OpenAI-wire when the anthropic SDK isn't installed. - Config hygiene: `_update_config_for_provider` (hermes_cli/auth.py) clears stale `api_key`/`api_mode` when switching to a built-in provider, so a previous MiniMax custom endpoint's credentials can't leak into a later OpenRouter session. - Truncation continuation: length-continuation and tool-call-truncation retry now cover `anthropic_messages` in addition to `chat_completions` and `bedrock_converse`. Reuses the existing `_build_assistant_message` path via `normalize_anthropic_response()` so the interim message shape is byte-identical to the non-truncated path. Tests: 6 new files, 42 test cases. Targeted run + tests/run_agent, tests/agent, tests/hermes_cli all pass (4554 passed). Synthesized from (credits preserved via Co-authored-by trailers): NousResearch#7410 @nocoo — URL detection helper NousResearch#7393 @keyuyuan — OAuth 5-site guard NousResearch#7367 @n-WN — OAuth guard (narrower cousin, kept comment) NousResearch#8636 @sgaofen — caching helper + native-vs-proxy layout split NousResearch#10954 @Only-Code-A — caching on anthropic_messages+Claude NousResearch#7648 @zhongyueming1121 — aux client anthropic_messages branch NousResearch#6096 @hansnow — /model switch clears stale api_mode NousResearch#9691 @TroyMitchell911 — anthropic_messages truncation continuation Closes: NousResearch#7366, NousResearch#8294 (third-party Anthropic identity + caching). Supersedes: NousResearch#7410, NousResearch#7367, NousResearch#7393, NousResearch#8636, NousResearch#10954, NousResearch#7648, NousResearch#6096, NousResearch#9691. Rejects: NousResearch#9621 (OpenAI-wire caching with incomplete blocklist — risky), NousResearch#7242 (superseded by NousResearch#9691, stale branch), NousResearch#8321 (targets smart_model_routing which was removed in NousResearch#12732). Co-authored-by: nocoo <nocoo@users.noreply.github.com> Co-authored-by: Keyu Yuan <leoyuan0099@gmail.com> Co-authored-by: Zoee <30841158+n-WN@users.noreply.github.com> Co-authored-by: sgaofen <135070653+sgaofen@users.noreply.github.com> Co-authored-by: Only-Code-A <bxzt2006@163.com> Co-authored-by: zhongyueming <mygamez@163.com> Co-authored-by: Xiaohan Li <hansnow@users.noreply.github.com> Co-authored-by: Troy Mitchell <i@troy-y.org>

@nocoo

…pport (NousResearch#12846) Third-party gateways that speak the native Anthropic protocol (MiniMax, Zhipu GLM, Alibaba DashScope, Kimi, LiteLLM proxies) now work end-to-end with the same feature set as direct api.anthropic.com callers. Synthesizes eight stale community PRs into one consolidated change. Five fixes: - URL detection: consolidate three inline `endswith("/anthropic")` checks in runtime_provider.py into the shared _detect_api_mode_for_url helper. Third-party /anthropic endpoints now auto-resolve to api_mode=anthropic_messages via one code path instead of three. - OAuth leak-guard: all five sites that assign `_is_anthropic_oauth` (__init__, switch_model, _try_refresh_anthropic_client_credentials, _swap_credential, _try_activate_fallback) now gate on `provider == "anthropic"` so a stale ANTHROPIC_TOKEN never trips Claude-Code identity injection on third-party endpoints. Previously only 2 of 5 sites were guarded. - Prompt caching: new method `_anthropic_prompt_cache_policy()` returns `(should_cache, use_native_layout)` per endpoint. Replaces three inline conditions and the `native_anthropic=(api_mode=='anthropic_messages')` call-site flag. Native Anthropic and third-party Anthropic gateways both get the native cache_control layout; OpenRouter gets envelope layout. Layout is persisted in `_primary_runtime` so fallback restoration preserves the per-endpoint choice. - Auxiliary client: `_try_custom_endpoint` honors `api_mode=anthropic_messages` and builds `AnthropicAuxiliaryClient` instead of silently downgrading to an OpenAI-wire client. Degrades gracefully to OpenAI-wire when the anthropic SDK isn't installed. - Config hygiene: `_update_config_for_provider` (hermes_cli/auth.py) clears stale `api_key`/`api_mode` when switching to a built-in provider, so a previous MiniMax custom endpoint's credentials can't leak into a later OpenRouter session. - Truncation continuation: length-continuation and tool-call-truncation retry now cover `anthropic_messages` in addition to `chat_completions` and `bedrock_converse`. Reuses the existing `_build_assistant_message` path via `normalize_anthropic_response()` so the interim message shape is byte-identical to the non-truncated path. Tests: 6 new files, 42 test cases. Targeted run + tests/run_agent, tests/agent, tests/hermes_cli all pass (4554 passed). Synthesized from (credits preserved via Co-authored-by trailers): NousResearch#7410 @nocoo — URL detection helper NousResearch#7393 @keyuyuan — OAuth 5-site guard NousResearch#7367 @n-WN — OAuth guard (narrower cousin, kept comment) NousResearch#8636 @sgaofen — caching helper + native-vs-proxy layout split NousResearch#10954 @Only-Code-A — caching on anthropic_messages+Claude NousResearch#7648 @zhongyueming1121 — aux client anthropic_messages branch NousResearch#6096 @hansnow — /model switch clears stale api_mode NousResearch#9691 @TroyMitchell911 — anthropic_messages truncation continuation Closes: NousResearch#7366, NousResearch#8294 (third-party Anthropic identity + caching). Supersedes: NousResearch#7410, NousResearch#7367, NousResearch#7393, NousResearch#8636, NousResearch#10954, NousResearch#7648, NousResearch#6096, NousResearch#9691. Rejects: NousResearch#9621 (OpenAI-wire caching with incomplete blocklist — risky), NousResearch#7242 (superseded by NousResearch#9691, stale branch), NousResearch#8321 (targets smart_model_routing which was removed in NousResearch#12732). Co-authored-by: nocoo <nocoo@users.noreply.github.com> Co-authored-by: Keyu Yuan <leoyuan0099@gmail.com> Co-authored-by: Zoee <30841158+n-WN@users.noreply.github.com> Co-authored-by: sgaofen <135070653+sgaofen@users.noreply.github.com> Co-authored-by: Only-Code-A <bxzt2006@163.com> Co-authored-by: zhongyueming <mygamez@163.com> Co-authored-by: Xiaohan Li <hansnow@users.noreply.github.com> Co-authored-by: Troy Mitchell <i@troy-y.org>

@nocoo

…pport (NousResearch#12846) Third-party gateways that speak the native Anthropic protocol (MiniMax, Zhipu GLM, Alibaba DashScope, Kimi, LiteLLM proxies) now work end-to-end with the same feature set as direct api.anthropic.com callers. Synthesizes eight stale community PRs into one consolidated change. Five fixes: - URL detection: consolidate three inline `endswith("/anthropic")` checks in runtime_provider.py into the shared _detect_api_mode_for_url helper. Third-party /anthropic endpoints now auto-resolve to api_mode=anthropic_messages via one code path instead of three. - OAuth leak-guard: all five sites that assign `_is_anthropic_oauth` (__init__, switch_model, _try_refresh_anthropic_client_credentials, _swap_credential, _try_activate_fallback) now gate on `provider == "anthropic"` so a stale ANTHROPIC_TOKEN never trips Claude-Code identity injection on third-party endpoints. Previously only 2 of 5 sites were guarded. - Prompt caching: new method `_anthropic_prompt_cache_policy()` returns `(should_cache, use_native_layout)` per endpoint. Replaces three inline conditions and the `native_anthropic=(api_mode=='anthropic_messages')` call-site flag. Native Anthropic and third-party Anthropic gateways both get the native cache_control layout; OpenRouter gets envelope layout. Layout is persisted in `_primary_runtime` so fallback restoration preserves the per-endpoint choice. - Auxiliary client: `_try_custom_endpoint` honors `api_mode=anthropic_messages` and builds `AnthropicAuxiliaryClient` instead of silently downgrading to an OpenAI-wire client. Degrades gracefully to OpenAI-wire when the anthropic SDK isn't installed. - Config hygiene: `_update_config_for_provider` (hermes_cli/auth.py) clears stale `api_key`/`api_mode` when switching to a built-in provider, so a previous MiniMax custom endpoint's credentials can't leak into a later OpenRouter session. - Truncation continuation: length-continuation and tool-call-truncation retry now cover `anthropic_messages` in addition to `chat_completions` and `bedrock_converse`. Reuses the existing `_build_assistant_message` path via `normalize_anthropic_response()` so the interim message shape is byte-identical to the non-truncated path. Tests: 6 new files, 42 test cases. Targeted run + tests/run_agent, tests/agent, tests/hermes_cli all pass (4554 passed). Synthesized from (credits preserved via Co-authored-by trailers): NousResearch#7410 @nocoo — URL detection helper NousResearch#7393 @keyuyuan — OAuth 5-site guard NousResearch#7367 @n-WN — OAuth guard (narrower cousin, kept comment) NousResearch#8636 @sgaofen — caching helper + native-vs-proxy layout split NousResearch#10954 @Only-Code-A — caching on anthropic_messages+Claude NousResearch#7648 @zhongyueming1121 — aux client anthropic_messages branch NousResearch#6096 @hansnow — /model switch clears stale api_mode NousResearch#9691 @TroyMitchell911 — anthropic_messages truncation continuation Closes: NousResearch#7366, NousResearch#8294 (third-party Anthropic identity + caching). Supersedes: NousResearch#7410, NousResearch#7367, NousResearch#7393, NousResearch#8636, NousResearch#10954, NousResearch#7648, NousResearch#6096, NousResearch#9691. Rejects: NousResearch#9621 (OpenAI-wire caching with incomplete blocklist — risky), NousResearch#7242 (superseded by NousResearch#9691, stale branch), NousResearch#8321 (targets smart_model_routing which was removed in NousResearch#12732). Co-authored-by: nocoo <nocoo@users.noreply.github.com> Co-authored-by: Keyu Yuan <leoyuan0099@gmail.com> Co-authored-by: Zoee <30841158+n-WN@users.noreply.github.com> Co-authored-by: sgaofen <135070653+sgaofen@users.noreply.github.com> Co-authored-by: Only-Code-A <bxzt2006@163.com> Co-authored-by: zhongyueming <mygamez@163.com> Co-authored-by: Xiaohan Li <hansnow@users.noreply.github.com> Co-authored-by: Troy Mitchell <i@troy-y.org>

BukeLy added 2 commits April 12, 2026 19:06

test(routing): add tests for api_mode passthrough in smart routing

7a47532

Verify that api_mode from cheap_model config takes priority over runtime-resolved value, falls back correctly when unset, and is reflected in the routing signature for cache keying.

teknium1 mentioned this pull request Apr 20, 2026

fix(anthropic): complete third-party Anthropic-compatible provider support #12846

Merged

teknium1 closed this Apr 20, 2026

BukeLy deleted the fix/smart-routing-api-mode branch April 20, 2026 06:28

rodboev mentioned this pull request Jun 2, 2026

[Bug]: Smart routing drops api_mode from cheap_model config #8515

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix(routing): pass api_mode from cheap_model config in smart routing#8321

fix(routing): pass api_mode from cheap_model config in smart routing#8321
BukeLy wants to merge 2 commits into
NousResearch:mainfrom
BukeLy:fix/smart-routing-api-mode

BukeLy commented Apr 12, 2026 •

edited

Loading

Uh oh!

BukeLy commented Apr 13, 2026

Uh oh!

teknium1 commented Apr 20, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

BukeLy commented Apr 12, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What does this PR do?

Related Issue

Type of Change

Changes Made

How to Test

Checklist

Code

Documentation & Housekeeping

Uh oh!

BukeLy commented Apr 13, 2026

Uh oh!

teknium1 commented Apr 20, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

BukeLy commented Apr 12, 2026 •

edited

Loading