Voxtral Small 24B 2507 is 8.5× cheaper per million tokens (blended). Qwen3.8 27B has the larger context window (1,000,000 vs 32,768 tokens, 30.5×).
Voxtral Small 24B 2507 vs Qwen3.8 27B
Live catalog fields. Best-in-row is highlighted.
Voxtral Small 24B 2507open
Mistral · mistralai__voxtral-small-24b-2507
Qwen3.8 27Bopen
Qwen · qwen__qwen3.8-27b
Identity
| Field | mistralai__voxtral-small-24b-2507 | qwen__qwen3.8-27b |
|---|---|---|
| Provider | Mistral | Qwen |
| Slug | mistralai__voxtral-small-24b-2507 | qwen__qwen3.8-27b |
| Status | live | live |
| Open weights | Yes | Yes |
| License | — | — |
| HuggingFace | mistralai/Voxtral-Small-24B-2507 | Qwen/Qwen3.8-27B |
Quality
| Field | mistralai__voxtral-small-24b-2507 | qwen__qwen3.8-27b |
|---|---|---|
| Quality index (0–100) | — | 87.4 |
Cost
| Field | mistralai__voxtral-small-24b-2507 | qwen__qwen3.8-27b |
|---|---|---|
| Input $ / 1M tokens | $0.10 | $0.42 |
| Cached input $ / 1M | $0.01 | $0.09 |
| Output $ / 1M tokens | $0.30 | $3.00 |
Context
| Field | mistralai__voxtral-small-24b-2507 | qwen__qwen3.8-27b |
|---|---|---|
| Max context tokens | 32,768 | 1,000,000 |
| Max output tokens | 26,214 | 131,072 |
Modalities
| Field | mistralai__voxtral-small-24b-2507 | qwen__qwen3.8-27b |
|---|---|---|
| Input | text, audio, file | text, image, video |
| Output | text | text |