Voxtral Small 24B 2507 vs Nova Lite vs Qwen Flash
Nova Lite comes out ahead, 77 to 68 and 55 on our weighted score, and it is the cheaper option too.
Mistral AIVoxtral Small 24B 2507
55/100- ECI—
- Price$0.10 / $0.30
- Context33K
- Our pick
AmazonNova Lite
77/100- ECI—
- Price$0.06 / $0.24
- Context300K
Alibaba (Qwen)Qwen Flash
68/100- ECI—
- Price$0.05 / $0.40
- Context1M
Nova Lite is our pick
Nova Lite is the better all-round choice, scoring 77/100 against Qwen Flash (68) and Voxtral Small 24B 2507 (55). It leads on price and inputs & features. Qwen Flash wins on context window. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
- CapabilityNot enough dataNo independent benchmark covers every model here yet
- Lowest priceNova LiteNova Lite $0.105 · Qwen Flash $0.138 · Voxtral Small 24B 2507 $0.15 per 1M tokens (3:1 blend)
- Longest contextQwen FlashQwen Flash 1,000,000 · Nova Lite 300,000 · Voxtral Small 24B 2507 32,768 tokens
- Widest inputsNova LiteVoxtral Small 24B 2507: Text, Audio · Nova Lite: Text, Images, PDFs, Video · Qwen Flash: Text
- Self-hostingVoxtral Small 24B 2507Publishes downloadable weights (Apache 2.0)
| Measure | Weight | Voxtral Small 24B 2507 | Nova Lite | Qwen Flash |
|---|---|---|---|---|
| Price | 50% | 89 | 96 | 91 |
| Inputs & features | 30% | 35 | 70 | 35 |
| Context window | 20% | 0 | 39 | 60 |
| Overall | 100% | 55/100 | 77/100 | 68/100 |
Left out because at least one model lacks the data: capability. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | Voxtral Small 24B 2507Mistral AI | Nova LiteAmazon | Qwen FlashAlibaba (Qwen) |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $0.10 | $0.06 | $0.05 (best) |
| Output | $0.30 | $0.24 (best) | $0.40 |
| Cached input | — | $0.015 | — |
| Blended (3:1) | $0.15 | $0.105 (best) | $0.138 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Official Mistral API | Official Amazon Bedrock API | Official Alibaba API |
| Limits | |||
| Context window | 32,768 tokens | 300,000 tokens | 1,000,000 tokens (best) |
| Max output | 32,768 tokens (best) | 10,000 tokens | 32,768 tokens (best) |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | No |
| PDFs | No | Yes | No |
| Audio | Yes | No | No |
| Video | No | Yes | No |
| Reasoning | No | No | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | No | No |
| Availability | |||
| Weights | OpenApache 2.0 | Proprietary | Proprietary |
| API model ID | voxtral-small-latest | amazon.nova-lite-v1:0 | qwen-flash |
| API providers | 7 (best) | 3 | 6 |
| Released | Jul 15, 2025 | Dec 3, 2024 | Jul 28, 2025 |
| Knowledge cutoff | — | Oct 2024 | Apr 2024 |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- Voxtral Small 24B 2507$1.60
- Nova Lite$1.08
- Qwen Flash$1.30
Which should you choose?
Which is better: Voxtral Small 24B 2507, Nova Lite or Qwen Flash?
Nova Lite is the better all-round choice, scoring 77/100 against Qwen Flash (68) and Voxtral Small 24B 2507 (55). It leads on price and inputs & features. Qwen Flash wins on context window. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
Which is cheaper, Voxtral Small 24B 2507, Nova Lite or Qwen Flash?
Nova Lite is cheaper at $0.06 input / $0.24 output per million tokens (official Amazon Bedrock API price). Qwen Flash costs $0.05 input / $0.40 output per million tokens (official Alibaba API price); Voxtral Small 24B 2507 costs $0.10 input / $0.30 output per million tokens (official Mistral API price). At a typical mix of three input tokens to one output token, that is $0.105 per million tokens for Nova Lite versus $0.138 for Qwen Flash (1.3× as much) and $0.15 for Voxtral Small 24B 2507 (1.4× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Voxtral Small 24B 2507 has not been scored yet, Nova Lite has not been scored yet and Qwen Flash has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for Voxtral Small 24B 2507, Nova Lite and Qwen Flash yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. All three support tool calling for agent workflows.
Which has the bigger context window?
Qwen Flash has the largest context window at 1,000,000 tokens, against 300,000 for Nova Lite and 32,768 for Voxtral Small 24B 2507. Maximum output per response: Voxtral Small 24B 2507 up to 32,768, Nova Lite up to 10,000, Qwen Flash up to 32,768 tokens.
Which can read images, PDFs, audio or video?
Voxtral Small 24B 2507 accepts text and audio; Nova Lite accepts text, images, PDFs and video; Qwen Flash accepts text. Nova Lite handles the widest range of inputs.
Are any of these open source?
Voxtral Small 24B 2507 publishes its weights (Apache 2.0) and can be self-hosted; Nova Lite and Qwen Flash is proprietary.
Which is newer?
Qwen Flash is the newest, released Jul 28, 2025. Voxtral Small 24B 2507 came out Jul 15, 2025; Nova Lite came out Dec 3, 2024. Knowledge cutoff: Nova Lite Oct 2024, Qwen Flash Apr 2024.
How do you decide the winner?
Each model gets a 0–100 score on capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Oct 4, 2026.
