Qwen3-VL 30B-A3B vs Codestral vs Llama 3.1 Nemotron 70B Instruct
Qwen3-VL 30B-A3B comes out ahead, 59 to 48 and 45 on our weighted score, and it is the cheaper option too.
- Our pick
Alibaba (Qwen)Qwen3-VL 30B-A3B
59/100- ECI—
- Price$0.20 / $0.80
- Context131K
Mistral AICodestral
48/100- ECI—
- Price$0.30 / $0.90
- Context256K
NVIDIALlama 3.1 Nemotron 70B Instruct
45/100- ECI—
- Price$0.478 / $0.504
- Context128K
Qwen3-VL 30B-A3B is our pick
Qwen3-VL 30B-A3B is the better all-round choice, scoring 59/100 against Codestral (48) and Llama 3.1 Nemotron 70B Instruct (45). It leads on price and inputs & features. Codestral 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 priceQwen3-VL 30B-A3BQwen3-VL 30B-A3B $0.35 · Codestral $0.45 · Llama 3.1 Nemotron 70B Instruct $0.485 per 1M tokens (3:1 blend)
- Longest contextCodestralCodestral 256,000 · Qwen3-VL 30B-A3B 131,072 · Llama 3.1 Nemotron 70B Instruct 128,000 tokens
- Widest inputsQwen3-VL 30B-A3BQwen3-VL 30B-A3B: Text, Images · Codestral: Text · Llama 3.1 Nemotron 70B Instruct: Text
- Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
| Measure | Weight | Qwen3-VL 30B-A3B | Codestral | Llama 3.1 Nemotron 70B Instruct |
|---|---|---|---|---|
| Price | 50% | 72 | 66 | 65 |
| Inputs & features | 30% | 60 | 25 | 25 |
| Context window | 20% | 24 | 36 | 24 |
| Overall | 100% | 59/100 | 48/100 | 45/100 |
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 | Qwen3-VL 30B-A3BAlibaba (Qwen) | CodestralMistral AI | Llama 3.1 Nemotron 70B InstructNVIDIA |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $0.20 (best) | $0.30 | $0.478 |
| Output | $0.80 | $0.90 | $0.504 (best) |
| Cached input | — | $0.03 | — |
| Blended (3:1) | $0.35 (best) | $0.45 | $0.485 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Official Alibaba API | Official Mistral API | Median of 2 providers |
| Limits | |||
| Context window | 131,072 tokens | 256,000 tokens (best) | 128,000 tokens |
| Max output | 32,768 tokens (best) | 4,096 tokens | 8,192 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | Yes | No | No |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | No | No |
| Reasoning | Yes | No | No |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | No | No |
| Availability | |||
| Weights | Open | Open | Open |
| API model ID | qwen3-vl-30b-a3b | codestral-latest | nvidia/llama-3.1-nemotron-70b-instruct |
| API providers | 1 | 3 (best) | 3 (best) |
| Released | Apr 2025 | May 29, 2024 | Apr 15, 2025 |
| Knowledge cutoff | Apr 2025 | Oct 2024 | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- Qwen3-VL 30B-A3B$3.60
- Codestral$4.80
- Llama 3.1 Nemotron 70B Instruct$5.79
Which should you choose?
Which is better: Qwen3-VL 30B-A3B, Codestral or Llama 3.1 Nemotron 70B Instruct?
Qwen3-VL 30B-A3B is the better all-round choice, scoring 59/100 against Codestral (48) and Llama 3.1 Nemotron 70B Instruct (45). It leads on price and inputs & features. Codestral 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, Qwen3-VL 30B-A3B, Codestral or Llama 3.1 Nemotron 70B Instruct?
Qwen3-VL 30B-A3B is cheaper at $0.20 input / $0.80 output per million tokens (official Alibaba API price). Codestral costs $0.30 input / $0.90 output per million tokens (official Mistral API price); Llama 3.1 Nemotron 70B Instruct costs $0.478 input / $0.504 output per million tokens (median across 2 API providers; free on Nvidia). At a typical mix of three input tokens to one output token, that is $0.35 per million tokens for Qwen3-VL 30B-A3B versus $0.45 for Codestral (1.3× as much) and $0.485 for Llama 3.1 Nemotron 70B Instruct (1.4× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Qwen3-VL 30B-A3B has not been scored yet, Codestral has not been scored yet and Llama 3.1 Nemotron 70B Instruct has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for Qwen3-VL 30B-A3B, Codestral and Llama 3.1 Nemotron 70B Instruct 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?
Codestral has the largest context window at 256,000 tokens, against 131,072 for Qwen3-VL 30B-A3B and 128,000 for Llama 3.1 Nemotron 70B Instruct. Maximum output per response: Qwen3-VL 30B-A3B up to 32,768, Codestral up to 4,096, Llama 3.1 Nemotron 70B Instruct up to 8,192 tokens.
Which can read images, PDFs, audio or video?
Qwen3-VL 30B-A3B accepts text and images; Codestral accepts text; Llama 3.1 Nemotron 70B Instruct accepts text. Qwen3-VL 30B-A3B handles the widest range of inputs.
Are any of these open source?
Yes, all three publish their weights, so you can self-host them.
Which is newer?
Llama 3.1 Nemotron 70B Instruct is the newest, released Apr 15, 2025. Qwen3-VL 30B-A3B came out Apr 2025; Codestral came out May 29, 2024. Knowledge cutoff: Qwen3-VL 30B-A3B Apr 2025, Codestral Oct 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.
