Mistral Small 4

Model overview
Mistral Small 4
Small 4 hybrid model from Mistral's official catalog.
general LLMreasoningcodingmultimodalvision-languagechatcode generationdocument understandinglocal inference
Representative version
Mistral Small 4
| Company | Mistral AI |
|---|---|
| Release | Not disclosed |
| Release type | Model release |
| Parameters | 119B |
| Architecture | Not disclosed |
| Context | Not disclosed |
| Input modalities | text / image / code |
| Output modalities | text / code |
| Tool use / function calling | Not disclosed |
| Structured output / reasoning | Confirmed: Reasoning |
| Vision / audio / video | Confirmed: Vision |
| Open weights / License | Weights available / Apache-2.0 License sourcemistral-small-4-phase-3-5 · 2026-07-15 |
| Vendor API | Not disclosed API source pending verification |
| OpenRouter | OpenRouter listed |
| Pricing | OpenRouter · 2026-07-15: $0.15 / $0.6 input / output, per 1M tokens |
| Benchmark | No reliable public source |
| Public signals | 1 runtime / access record / 1 activity snapshot |
| Use cases | assistant chat / code generation / document understanding / local inference |
| Claim evidence | Release eventmistralai/Mistral-Small-4-119B-2603 Hugging Face model cardMistral Small 4 Model specs / weightsMistral models overviewMistral Small 4 · Mistral Small 4 |
OpenRouter
Pricing / context / availability
| OpenRouter | OpenRouter listed |
|---|---|
| OpenRouter listing | mistralai/mistral-small-2603 |
| Mapped version | Mistral Small 4 mistral-small-4-phase-3-5 |
| Pricing | OpenRouter · 2026-07-15: $0.15 / $0.6 input / output, per 1M tokens |
| Context | 262144 |
| Speed/Latency | No reliable public source |
| Snapshot time | 2026-07-15T20:11:41.551Z |
| Source | View source |
Public activity counters
Cumulative Hugging Face model-card downloads / likes / spaces
| Platform | Entry | Downloads | Likes | Spaces | Time | Source |
|---|---|---|---|---|---|---|
| huggingface | mistralai/Mistral-Small-4-119B-2603 | 217,363 | 409 | 23 | 2026-07-15T20:11:41.551Z | View source |
Operations
Runtime / access signals
| Time | Signal | Original text | Source |
|---|---|---|---|
| 06/24/2026, 05:54 PM UTC | Anyone running this on 2×H200 with vLLM? Hangs after "Fp8 MoE backend" | Hi! Has anyone successfully served Mistral-Small-4-119B-2603 on 2×H200 with vLLM? I'm stuck at startup. What happens: The weights load fine to... | Source |
Evidence gaps
Public display boundary
Release date not disclosedNo reliable benchmark sourceNo linkable real reviewSpeed/latency not disclosed