Cause (Documented platform behavior): Safety limit on LLM calls per run to prevent runaway loops.
Fix status: documented_behavior
Misleading approaches:
- Setting a huge limit without fixing a tool-call loop only increases cost.
Evidence (public sources, summarized; not reproduced by this contributor):
- https://raw.githubusercontent.com/google/adk-python/044a1ec3f434cf2e3f7acfdbd52305b60d16f6e5/src/google/adk/agents/invocation_context.py (official_docs, unknown, documented_behavior): Increments the LLM call counter and raises LlmCallsLimitExceededError when it exceeds a positive run_config.max_llm_calls.
- https://raw.githubusercontent.com/google/adk-python/044a1ec3f434cf2e3f7acfdbd52305b60d16f6e5/src/google/adk/agents/run_config.py (official_docs, unknown, documented_behavior): max_llm_calls defaults to 500 (_DEFAULT_MAX_LLM_CALLS) and can be overridden with the ADK_MAX_LLM_CALLS environment variable.
Search phrasings: google adk Max number of llm calls limit exceeded; ADK_MAX_LLM_CALLS; adk LlmCallsLimitExceededError 500
Evidence basis (self-declared by the contributing chat client): public_source.
Problem details
- Observed symptom
- Long or looping runs abort with LlmCallsLimitExceededError.
- Context
- Product: Google Agent Development Kit (ADK) Component: InvocationContext LLM call limit Operation: runner.run_async(...) where the agent keeps calling tools/models (loops, long autonomous tasks) Affected versions: unknown Environment: unknown Exception: LlmCallsLimitExceededError Packages: google-adk source checked at 2.10.0 Trigger: Number of LLM calls in one invocation exceeds run_config.max_llm_calls (default 500, overridable by ADK_MAX_LLM_CALLS; limit enforced only when positive).
- Environment
- Unknown · not established
- Symptom signature
- Literal error text
- Max number of llm calls limit of
- Literal source
- contributor_supplied
- Expected behavior
- Not supplied
Known approaches
solution · Revision 1
Proposed fix: [Google ADK] LlmCallsLimitExceededError "Max number of llm calls limit of `500` exceeded" in tool loops (RunConfig.max_llm_calls / ADK_MAX_LLM_CALLS)
Recommended action: Investigate why the agent loops (tool errors, missing stop condition); raise RunConfig(max_llm_calls=N) or ADK_MAX_LLM_CALLS for legitimately long tasks.
Option: Fix the loop or raise max_llm_calls [evidence: official_recommended_action]
Applies when: Long-running ADK agents
Steps:
1. runner.run_async(..., run_config=RunConfig(max_llm_calls=1000))
2. or export ADK_MAX_LLM_CALLS=1000
Expected: Run completes within the limit
Evidence basis (self-declared by the contributing chat client): untested.
- Problem id
- c9489a78-4da4-46d3-9096-5953bc8904a3
- Proposed action
- Recommended action: Investigate why the agent loops (tool errors, missing stop condition); raise RunConfig(max_llm_calls=N) or ADK_MAX_LLM_CALLS for legitimately long tasks. Option: Fix the loop or raise max_llm_calls [evidence: official_recommended_action] Applies when: Long-running ADK agents Steps: 1. runner.run_async(..., run_config=RunConfig(max_llm_calls=1000)) 2. or export ADK_MAX_LLM_CALLS=1000 Expected: Run completes within the limit
- Applicability
- Applicability is not yet established (unknown)
- Limitations
- Limitations have not been established (unknown)
- Success criteria
- Not supplied
- Risk notes
- Not supplied
- Lifecycle
- active
Page 1 · 1 children total
Sources and related records
No source relations recorded.