Vision Large vs Hermes 3 405B Instruct vs Codestral-22B-v0.1
Vision Large comes out ahead, 84 to 10 and 0 on our weighted score, though Hermes 3 405B Instruct is 11× cheaper per token.
- Our pick
VisparkVision Large
84/100- ECI—
- Price$7.37 / $22.11
- Context1M
Nous ResearchHermes 3 405B Instruct
10/100- ECI—
- Price$1.00 / $1.00
- Context131K
Mistral AICodestral-22B-v0.1
0/100- ECI—
- Price—
- Context33K
Vision Large is our pick
Vision Large is the better all-round choice, scoring 84/100 against Hermes 3 405B Instruct (10) and Codestral-22B-v0.1 (0). 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 priceHermes 3 405B InstructHermes 3 405B Instruct $1.00 · Vision Large $11.05 per 1M tokens (3:1 blend) · Codestral-22B-v0.1 unpriced
- Longest contextVision LargeVision Large 1,000,000 · Hermes 3 405B Instruct 131,072 · Codestral-22B-v0.1 32,768 tokens
- Widest inputsVision LargeVision Large: Text, Images, PDFs, Audio, Video · Hermes 3 405B Instruct: Text · Codestral-22B-v0.1: Text
- Self-hostingHermes 3 405B Instruct and Codestral-22B-v0.1Publishes downloadable weights (Mistral AI Non-Production License)
| Measure | Weight | Vision Large | Hermes 3 405B Instruct | Codestral-22B-v0.1 |
|---|---|---|---|---|
| Inputs & features | 60% | 100 | 0 | 0 |
| Context window | 40% | 60 | 25 | 0 |
| Overall | 100% | 84/100 | 10/100 | 0/100 |
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 | Vision LargeVispark | Hermes 3 405B InstructNous Research | Codestral-22B-v0.1Mistral AI |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $7.37 | $1.00 (best) | — |
| Output | $22.11 | $1.00 (best) | — |
| Cached input | — | — | — |
| Blended (3:1) | $11.05 | $1.00 (best) | — |
| Long-context rate | Same rate | Same rate | — |
| Price source | Official Vispark API | Median of 5 providers | — |
| Limits | |||
| Context window | 1,000,000 tokens (best) | 131,072 tokens | 32,768 tokens |
| Max output | 65,536 tokens (best) | 16,384 tokens | 8,192 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | Yes | No | No |
| PDFs | Yes | No | No |
| Audio | Yes | No | No |
| Video | Yes | No | No |
| Reasoning | Yes | No | No |
| Tool calling | Yes | No | No |
| Structured output | Yes | No | No |
| Availability | |||
| Weights | Proprietary | Open | OpenMistral AI Non-Production License |
| API model ID | vispark/vision-large | — | — |
| API providers | 1 | 5 (best) | — |
| Released | May 15, 2024 | Aug 16, 2024 | May 29, 2024 |
| Knowledge cutoff | — | Dec 31, 2023 | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- Vision Large$117.92
- Hermes 3 405B Instruct$12.00
- Codestral-22B-v0.1—
Which should you choose?
Which is better: Vision Large, Hermes 3 405B Instruct or Codestral-22B-v0.1?
Vision Large is the better all-round choice, scoring 84/100 against Hermes 3 405B Instruct (10) and Codestral-22B-v0.1 (0). 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, Vision Large, Hermes 3 405B Instruct or Codestral-22B-v0.1?
Hermes 3 405B Instruct is cheaper at $1.00 input / $1.00 output per million tokens (median across 5 API providers). Vision Large costs $7.37 input / $22.11 output per million tokens (official Vispark API price). At a typical mix of three input tokens to one output token, that is $1.00 per million tokens for Hermes 3 405B Instruct versus $11.05 for Vision Large (11× as much). Codestral-22B-v0.1 has no published per-token price.
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Vision Large has not been scored yet, Hermes 3 405B Instruct has not been scored yet and Codestral-22B-v0.1 has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for Vision Large, Hermes 3 405B Instruct and Codestral-22B-v0.1 yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. Note that Hermes 3 405B Instruct and Codestral-22B-v0.1 does not support tool calling, which most coding agents need.
Which has the bigger context window?
Vision Large has the largest context window at 1,000,000 tokens, against 131,072 for Hermes 3 405B Instruct and 32,768 for Codestral-22B-v0.1. Maximum output per response: Vision Large up to 65,536, Hermes 3 405B Instruct up to 16,384, Codestral-22B-v0.1 up to 8,192 tokens.
Which can read images, PDFs, audio or video?
Vision Large accepts text, images, PDFs, audio and video; Hermes 3 405B Instruct accepts text; Codestral-22B-v0.1 accepts text. Vision Large handles the widest range of inputs.
Are any of these open source?
Hermes 3 405B Instruct and Codestral-22B-v0.1 publishes its weights (Mistral AI Non-Production License) and can be self-hosted; Vision Large is proprietary.
Which is newer?
Hermes 3 405B Instruct is the newest, released Aug 16, 2024. Codestral-22B-v0.1 came out May 29, 2024; Vision Large came out May 15, 2024. Knowledge cutoff: Hermes 3 405B Instruct Dec 31, 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 Nov 2, 2026.
