Codestral vs Llama-3.2-3B vs Vision Small
Vision Small comes out ahead, 84 to 29 and 10 on our weighted score.
Mistral AICodestral
29/100- ECI—
- Price$0.30 / $0.90
- Context256K
MetaLlama-3.2-3B
10/100- ECI—
- Price$0.10 / $0.335
- Context131K
- Our pick
VisparkVision Small
84/100- ECI—
- Price—
- Context1M
Vision Small is our pick
Vision Small is the better all-round choice, scoring 84/100 against Codestral (29) and Llama-3.2-3B (10). It leads on inputs & features and context window. The score weighs inputs & features 60%, context window 40%. 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 priceLlama-3.2-3BLlama-3.2-3B $0.159 · Codestral $0.45 per 1M tokens (3:1 blend) · Vision Small unpriced
- Longest contextVision SmallVision Small 1,000,000 · Codestral 256,000 · Llama-3.2-3B 131,072 tokens
- Widest inputsVision SmallCodestral: Text · Llama-3.2-3B: Text · Vision Small: Text, Images, PDFs, Audio, Video
- Self-hostingCodestral and Llama-3.2-3BPublishes downloadable weights (Llama 3.2 Community License)
| Measure | Weight | Codestral | Llama-3.2-3B | Vision Small |
|---|---|---|---|---|
| Inputs & features | 60% | 25 | 0 | 100 |
| Context window | 40% | 36 | 24 | 60 |
| Overall | 100% | 29/100 | 10/100 | 84/100 |
Left out because at least one model lacks the data: capability and price. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | CodestralMistral AI | Llama-3.2-3BMeta | Vision SmallVispark |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $0.30 | $0.10 (best) | — |
| Output | $0.90 | $0.335 (best) | — |
| Cached input | $0.03 | — | — |
| Blended (3:1) | $0.45 | $0.159 (best) | — |
| Long-context rate | Same rate | Same rate | — |
| Price source | Official Mistral API | Median of 3 providers | — |
| Limits | |||
| Context window | 256,000 tokens | 131,072 tokens | 1,000,000 tokens (best) |
| Max output | 4,096 tokens | 8,192 tokens | 65,536 tokens (best) |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | No | Yes |
| PDFs | No | No | Yes |
| Audio | No | No | Yes |
| Video | No | No | Yes |
| Reasoning | No | No | Yes |
| Tool calling | Yes | No | Yes |
| Structured output | No | No | Yes |
| Availability | |||
| Weights | Open | OpenLlama 3.2 Community License | Proprietary |
| API model ID | codestral-latest | — | — |
| API providers | 3 | 3 | — |
| Released | May 29, 2024 | Sep 25, 2024 | May 15, 2024 |
| Knowledge cutoff | Oct 2024 | Dec 2023 | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- Codestral$4.80
- Llama-3.2-3B$1.67
- Vision Small—
Which should you choose?
Which is better: Codestral, Llama-3.2-3B or Vision Small?
Vision Small is the better all-round choice, scoring 84/100 against Codestral (29) and Llama-3.2-3B (10). It leads on inputs & features and context window. The score weighs inputs & features 60%, context window 40%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
Which is cheaper, Codestral, Llama-3.2-3B or Vision Small?
Llama-3.2-3B is cheaper at $0.10 input / $0.335 output per million tokens (median across 3 API providers). Codestral costs $0.30 input / $0.90 output per million tokens (official Mistral API price). At a typical mix of three input tokens to one output token, that is $0.159 per million tokens for Llama-3.2-3B versus $0.45 for Codestral (2.8× as much). Vision Small has no published per-token price.
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Codestral has not been scored yet, Llama-3.2-3B has not been scored yet and Vision Small has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for Codestral, Llama-3.2-3B and Vision Small yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. Note that Llama-3.2-3B does not support tool calling, which most coding agents need.
Which has the bigger context window?
Vision Small has the largest context window at 1,000,000 tokens, against 256,000 for Codestral and 131,072 for Llama-3.2-3B. Maximum output per response: Codestral up to 4,096, Llama-3.2-3B up to 8,192, Vision Small up to 65,536 tokens.
Which can read images, PDFs, audio or video?
Codestral accepts text; Llama-3.2-3B accepts text; Vision Small accepts text, images, PDFs, audio and video. Vision Small handles the widest range of inputs.
Are any of these open source?
Codestral and Llama-3.2-3B publishes its weights (Llama 3.2 Community License) and can be self-hosted; Vision Small is proprietary.
Which is newer?
Llama-3.2-3B is the newest, released Sep 25, 2024. Codestral came out May 29, 2024; Vision Small came out May 15, 2024. Knowledge cutoff: Codestral Oct 2024, Llama-3.2-3B Dec 2023.
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.
