mirror of
https://github.com/BerriAI/litellm.git
synced 2026-09-28 01:32:17 +00:00
07a360bb1b
Adaptive-thinking models (Claude 4.7+) use thinking.type == "adaptive"
rather than "enabled", so is_thinking_enabled() — which only matches
"enabled" or reasoning_effort — reports False for them. On the
response_format → synthetic-tool fallback path, the
"if not is_thinking_enabled" guard then synthesizes a forced
tool_choice = {"name": "json_tool_call", "type": "tool"} while
"thinking": {"type": "adaptive"} is still in the optional_params, and
Anthropic / Bedrock reject the request with:
Thinking may not be enabled when tool_choice forces tool use.
This is unreachable for the in-tree Claude 4.6/4.7 models today because
they all match the response_format output_format whitelist a few lines
earlier, but it bites any deployment that pushes those models off the
native output_format path (e.g. yaml override of
supports_response_schema, or future adaptive models that aren't yet on
the whitelist). A parallel symptom in another SDK is documented in
vercel/ai#14773.
Add a carve-out using the existing _is_adaptive_thinking_model() helper
so the forced tool_choice is skipped for adaptive-thinking models. The
synthetic JSON tool itself is still added, and json_mode is still set,
so the response path is unchanged — the model can still emit JSON via
the tool when it chooses to.
Two unit tests:
- test_adaptive_thinking_model_response_format_fallback_no_forced_tool_choice
pins the new behaviour
- test_non_adaptive_model_response_format_fallback_still_forces_tool_choice
guards against silently widening the carve-out to non-adaptive models
|
||
|---|---|---|
| .. | ||
| agent_tests | ||
| audio_tests | ||
| basic_proxy_startup_tests | ||
| batches_tests | ||
| benchmarks | ||
| code_coverage_tests | ||
| documentation_tests | ||
| enterprise | ||
| guardrails_tests | ||
| image_gen_tests | ||
| litellm | ||
| litellm-proxy-extras | ||
| litellm_core_utils | ||
| litellm_utils_tests | ||
| llm_responses_api_testing | ||
| llm_translation | ||
| load_tests | ||
| local_testing | ||
| logging_callback_tests | ||
| mcp_tests | ||
| multi_instance_e2e_tests | ||
| ocr_tests | ||
| old_proxy_tests/tests | ||
| openai_endpoints_tests | ||
| otel_tests | ||
| pass_through_tests | ||
| pass_through_unit_tests | ||
| proxy_admin_ui_tests | ||
| proxy_e2e_anthropic_messages_tests | ||
| proxy_security_tests | ||
| proxy_unit_tests | ||
| router_unit_tests | ||
| scim_tests | ||
| search_tests | ||
| spend_tracking_tests | ||
| store_model_in_db_tests | ||
| test_litellm | ||
| unified_google_tests | ||
| vector_store_tests | ||
| windows_tests | ||
| __init__.py | ||
| _flush_vcr_cache.py | ||
| _vcr_conftest_common.py | ||
| _vcr_redis_persister.py | ||
| eval_swe_bench.py | ||
| gettysburg.wav | ||
| large_text.py | ||
| openai_batch_completions.jsonl | ||
| README.MD | ||
| test_budget_management.py | ||
| test_callbacks_on_proxy.py | ||
| test_config.py | ||
| test_debug_warning.py | ||
| test_default_encoding_non_root.py | ||
| test_end_users.py | ||
| test_entrypoint.py | ||
| test_fallbacks.py | ||
| test_gpt5_azure_temperature_support.py | ||
| test_health.py | ||
| test_keys.py | ||
| test_litellm_proxy_responses_config.py | ||
| test_logging.conf | ||
| test_models.py | ||
| test_new_vector_store_endpoints.py | ||
| test_openai_endpoints.py | ||
| test_organizations.py | ||
| test_otel_thread_leak.py | ||
| test_passthrough_endpoints.py | ||
| test_presidio_latency.py | ||
| test_proxy_server_non_root.py | ||
| test_ratelimit.py | ||
| test_resource_cleanup.py | ||
| test_service_logger_otel.py | ||
| test_spend_logs.py | ||
| test_team.py | ||
| test_team_logging.py | ||
| test_team_members.py | ||
| test_users.py | ||
In total litellm runs 1000+ tests
[02/20/2025] Update:
To make it easier to contribute and map what behavior is tested,
we've started mapping the litellm directory in tests/test_litellm
This folder can only run mock tests.