Create gpt-5-mini.toml #140
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
neilberkman
added a commit
to neilberkman/models.dev
that referenced
this pull request
Nov 1, 2025
- Update schema to accept boolean OR object for tool_call - Add streaming and coerces_types optional fields - Update meta.llama3-3-70b-instruct-v1:0 as example - Documents Bedrock streaming limitation and type coercion Addresses provider-specific tool calling limitations documented in: - AWS re:Post (streaming, type coercion issues) - langchain-aws anomalyco#140, anomalyco#354 - pydantic-ai #1649 100% backward compatible at TOML level - existing tool_call = true/false still valid
This was referenced Nov 1, 2025
neilberkman
added a commit
to neilberkman/models.dev
that referenced
this pull request
Nov 1, 2025
- Update schema to accept boolean OR object for tool_call - Add streaming, parallel, and coerces_types optional fields - Update meta.llama3-3-70b-instruct-v1:0 as example - Documents Bedrock streaming limitation and type coercion Addresses provider-specific tool calling limitations: - Streaming incompatibility (AWS Bedrock non-Anthropic models) - Type coercion issues (Bedrock integer handling) - Parallel tool calling tracking (addresses anomalyco#202) Evidence: - AWS re:Post: streaming errors, integer coercion - langchain-aws anomalyco#140, anomalyco#354: streaming validation issues - pydantic-ai #1649: response format problems - req_llm: agentjido/req_llm#163 100% backward compatible at TOML level - existing tool_call = true/false still valid
neilberkman
added a commit
to neilberkman/models.dev
that referenced
this pull request
Nov 1, 2025
- Update schema to accept boolean OR object for tool_call - Add streaming, parallel, and coerces_types optional fields - Update meta.llama3-3-70b-instruct-v1:0 as example - Documents Bedrock streaming limitation and type coercion Default semantics (do the least harm): - streaming: undefined → true (matches current tool_call=true assumption) - parallel: undefined → false (safer, not universally supported) - coerces_types: undefined → false (most models work correctly) Addresses provider-specific tool calling limitations: - Streaming incompatibility (AWS Bedrock non-Anthropic models) - Type coercion issues (Bedrock integer handling) - Parallel tool calling tracking (addresses anomalyco#202) Evidence: - AWS re:Post: streaming errors, integer coercion - langchain-aws anomalyco#140, anomalyco#354: streaming validation issues - pydantic-ai #1649: response format problems - req_llm: agentjido/req_llm#163 100% backward compatible at TOML level - existing tool_call = true/false still valid
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary:
• Introduces new model entry GPT-5-mini with 128k context and 128k output limits
• Enables text+image input modality and text output
• Flags advanced capabilities: reasoning mode, tool calls, temperature control, attachments
• Sets metadata: release_date 2025-08-13, last_updated 2025-08-13, knowledge cutoff 2024-06
• Explicitly marks weights as closed (open_weights = false)
Details:
• Model identifier: GPT-5-mini
• Capabilities: attachment handling, reasoning enhancements, external tool invocation, adjustable temperature
• Modalities: input (text, image), output (text)
• Limits: 128,000 token context window, 128,000 token max generation