docs: provider vision-input notes + accuracy fixes - #303
Merged
Conversation
HuggingFace vision-capable models accept image input via the router endpoint (https://router.huggingface.co/v1), enabled by conductor-oss/conductor#1245. Support is model-dependent. Added to both provider docs. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Extend the vision-input provider notes beyond HuggingFace to the other newly media-enabled providers, each referencing its conductor-oss fix: Anthropic (#1238), Gemini (#1241), Grok/Perplexity (#1243). Support is model-dependent. Applied to both provider docs. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
ling-senpeng13
changed the title
…lama, Bedrock Complete the per-provider vision-input coverage: these providers already forward media today (via OpenAIResponsesChatModel / Spring AI stock chat models), so note image input as natively supported (model-dependent), no PR reference needed. Every vision-capable provider in the docs now has a Vision (image input) note; Cohere/DeepSeek/StabilityAI are excluded (no vision chat models). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Correcting an earlier omission: Cohere IS vision-capable (e.g. command-a-vision-07-2025) and its v2 chat API accepts image_url content parts. Enabled in conductor-ai by conductor-oss/conductor#1246. Support is model-dependent. Added to both provider docs. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
DeepSeek is not a conductor-ai provider (no implementation, no registered alias), so `deepseek/deepseek-chat` would not resolve, and DeepSeek's API is text-only anyway. Remove the DeepSeek section and summary row from providers.md (ai-models.md never listed it). Re-add if/when a DeepSeek provider lands in conductor-ai. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
ling-senpeng13
changed the title
v1r3n
approved these changes
Jul 10, 2026
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
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.
Provider-docs accuracy pass for
docs/providers.mdanddocs/ai-models.md.1. Vision (image input) notes per provider
Adds a Vision (image input) note to every vision-capable provider so the docs uniformly state which providers accept image input.
Native (forward media today): OpenAI, Azure OpenAI, Mistral (Pixtral), Ollama (llava/llama3.2-vision), Bedrock — model-dependent, no PR needed.
Enabled by a conductor-oss fix:
command-a-vision-07-2025)Every note states support is model-dependent.
2. Accuracy fix — remove DeepSeek
providers.mdlisted a DeepSeek section + summary row, but DeepSeek has no conductor-ai provider (no implementation, no registereddeepseek/alias) and its API is text-only.ai-models.mdnever listed it. Removed the DeepSeek entry until a provider actually lands.Excluded (correctly): StabilityAI (image-generation only, no chat model).
🤖 Generated with Claude Code