The request names the LLaVA-NeXT (LLaVA v1.6) vision-language model built on Mistral, and the closest match is Replicate's hosted yorickvp/llava-v1.6-mistral-7b action 14. The tiny-random variant is not shown in any source, so the listed products are alternatives that run image-plus-text Mistral-family models.
A question back: Do you need the tiny-random test checkpoint specifically, or any working LLaVA-NeXT Mistral model?
- 1
- 2
Sources
- 1Replicate - Run AI with an APIreplicate.com · read 2026-10-29
- 2OpenRouteropenrouter.ai · read 2026-10-29
- 4yorickvp/llava-v1.6-mistral-7b: declares "yorickvp/llava-v1.6-mistral-7b" (the record we read and signed)unfragile.ai · read 2026-10-24
- 5Mistral: Mistral Small 4 (batch): declares "text+image->text" (the record we read and signed)unfragile.ai · read 2026-10-24
For agents: the 10 actions behind this answer
yorickvp/llava-v1.6-mistral-7b · yorickvp/llava-v1.6-mistral-7b verified and signed
LLaVA v1.6: Large Language and Vision Assistant (Mistral-7B)
mistralai/mistral-7b-v0.1 · mistralai/mistral-7b-v0.1 verified and signed
A 7 billion parameter language model from Mistral.
antoinelyset/openhermes-2-mistral-7b-awq · antoinelyset/openhermes-2-mistral-7b-awq verified and signed
nateraw/mistral-7b-openorca · nateraw/mistral-7b-openorca verified and signed
Mistral-7B-v0.1 fine tuned for chat with the OpenOrca dataset.
text+image->text · Mistral: Mistral Small 4 (batch) verified and signed
Takes text, image; returns text.
antoinelyset/openhermes-2.5-mistral-7b · antoinelyset/openhermes-2.5-mistral-7b verified and signed
text+image->text · Mistral: Mistral Small 3.1 24B verified and signed
Takes text, image; returns text.
lucataco/dolphin-2.2.1-mistral-7b · lucataco/dolphin-2.2.1-mistral-7b verified and signed
Mistral-7B-v0.1 fine tuned for chat with the Dolphin dataset (an open-source implementation of Microsoft's Orca)
text->text · Mistral: Mistral Small 3 verified and signed
Takes text; returns text.
text+image->text · Mistral: Ministral 3 3B 2512 verified and signed
Takes text, image; returns text.
Labels are the instruments' verdict and are never changed by ranking. Agents and robots get this same answer as JSON at /api/v1/find?q=…&answer=1 and through the MCP server.