Llama 4 Maverick 17B Instruct vs Qwen3 235B-A22B Instruct 2507 vs DeepSeek-V3
Too close to call on our weighted score (Qwen3 235B-A22B Instruct 2507 58, Llama 4 Maverick 17B Instruct 58, DeepSeek-V3 50). The right pick depends on what you value most.
MetaLlama 4 Maverick 17B Instruct
58/100- ECI132.2
- Price$0.321 / $0.91
- Context1M
Alibaba (Qwen)Qwen3 235B-A22B Instruct 2507
58/100- ECI138.9
- Price$0.15 / $0.75
- Context262K
DeepSeekDeepSeek-V3
50/100- ECI132.3
- Price$0.32 / $1.10
- Context131K
Too close to call
It is close. Our weighted score puts them within a point (Qwen3 235B-A22B Instruct 2507 58/100, Llama 4 Maverick 17B Instruct 58/100, DeepSeek-V3 50/100), so choose by what matters most for your work: Qwen3 235B-A22B Instruct 2507 for raw capability and Llama 4 Maverick 17B Instruct for long inputs. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
- CapabilityQwen3 235B-A22B Instruct 2507Capabilities Index (ECI): Qwen3 235B-A22B Instruct 2507 138.9 · DeepSeek-V3 132.3 · Llama 4 Maverick 17B Instruct 132.2
- Lowest priceQwen3 235B-A22B Instruct 2507Qwen3 235B-A22B Instruct 2507 $0.30 · Llama 4 Maverick 17B Instruct $0.468 · DeepSeek-V3 $0.515 per 1M tokens (3:1 blend)
- Longest contextLlama 4 Maverick 17B InstructLlama 4 Maverick 17B Instruct 1,000,000 · Qwen3 235B-A22B Instruct 2507 262,144 · DeepSeek-V3 131,072 tokens
- Widest inputsLlama 4 Maverick 17B InstructLlama 4 Maverick 17B Instruct: Text, Images · Qwen3 235B-A22B Instruct 2507: Text · DeepSeek-V3: Text
- Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
| Measure | Weight | Llama 4 Maverick 17B Instruct | Qwen3 235B-A22B Instruct 2507 | DeepSeek-V3 |
|---|---|---|---|---|
| CapabilityCapabilities Index (ECI) | 50% | 56 | 64 | 56 |
| Price | 25% | 66 | 75 | 64 |
| Inputs & features | 15% | 50 | 25 | 25 |
| Context window | 10% | 60 | 37 | 24 |
| Overall | 100% | 58/100 | 58/100 | 50/100 |
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | Llama 4 Maverick 17B InstructMeta | Qwen3 235B-A22B Instruct 2507Alibaba (Qwen) | DeepSeek-V3DeepSeek |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 132.2 | 138.9 (best) | 132.3 |
| ECI rank | #122 of 148 | #105 of 148 (best) | #121 of 148 |
| GPQA DiamondGraduate-level science questions | 67.0% (best) | — | 56.5% |
| OTIS Mock AIME 2024–2025Competition mathematics | 20.6% (best) | — | 15.8% |
| Price per million tokens | |||
| Input | $0.321 | $0.15 (best) | $0.32 |
| Output | $0.91 | $0.75 (best) | $1.10 |
| Cached input | — | — | — |
| Blended (3:1) | $0.468 | $0.30 (best) | $0.515 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Median of 6 providers | Median of 11 providers | Median of 5 providers |
| Limits | |||
| Context window | 1,000,000 tokens (best) | 262,144 tokens | 131,072 tokens |
| Max output | 16,384 tokens (best) | 16,384 tokens (best) | 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 | No | No | No |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | No | No |
| Availability | |||
| Weights | Open | OpenApache 2.0 | OpenDeepSeek Model License |
| API model ID | — | — | — |
| API providers | 6 | 11 (best) | 5 |
| Released | Apr 5, 2025 | Jul 21, 2025 | Dec 26, 2024 |
| Knowledge cutoff | Aug 2024 | — | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- Llama 4 Maverick 17B Instruct$5.03
- Qwen3 235B-A22B Instruct 2507$3.00
- DeepSeek-V3$5.40
Which should you choose?
Which is better: Llama 4 Maverick 17B Instruct, Qwen3 235B-A22B Instruct 2507 or DeepSeek-V3?
It is close. Our weighted score puts them within a point (Qwen3 235B-A22B Instruct 2507 58/100, Llama 4 Maverick 17B Instruct 58/100, DeepSeek-V3 50/100), so choose by what matters most for your work: Qwen3 235B-A22B Instruct 2507 for raw capability and Llama 4 Maverick 17B Instruct for long inputs. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
Which is cheaper, Llama 4 Maverick 17B Instruct, Qwen3 235B-A22B Instruct 2507 or DeepSeek-V3?
Qwen3 235B-A22B Instruct 2507 is cheaper at $0.15 input / $0.75 output per million tokens (median across 11 API providers). Llama 4 Maverick 17B Instruct costs $0.321 input / $0.91 output per million tokens (median across 6 API providers); DeepSeek-V3 costs $0.32 input / $1.10 output per million tokens (median across 5 API providers). At a typical mix of three input tokens to one output token, that is $0.30 per million tokens for Qwen3 235B-A22B Instruct 2507 versus $0.468 for Llama 4 Maverick 17B Instruct (1.6× as much) and $0.515 for DeepSeek-V3 (1.7× as much).
Which scores higher on benchmarks?
Qwen3 235B-A22B Instruct 2507 scores higher on the Capabilities Index (ECI): Qwen3 235B-A22B Instruct 2507 138.9 (#105 of 148), DeepSeek-V3 132.3 (#121 of 148) and Llama 4 Maverick 17B Instruct 132.2 (#122 of 148). Their confidence ranges do not overlap (135.8–140.6 vs 127.5–135.5), so the gap is a real one.
Which is better for coding?
There are no published SWE-bench Verified results for Llama 4 Maverick 17B Instruct, Qwen3 235B-A22B Instruct 2507 and DeepSeek-V3 yet, so there is no like-for-like coding score. On overall capability, Qwen3 235B-A22B Instruct 2507 leads, which tends to carry over to coding, but test on your own codebase. All three support tool calling for agent workflows.
Which has the bigger context window?
Llama 4 Maverick 17B Instruct has the largest context window at 1,000,000 tokens, against 262,144 for Qwen3 235B-A22B Instruct 2507 and 131,072 for DeepSeek-V3. Maximum output per response: Llama 4 Maverick 17B Instruct up to 16,384, Qwen3 235B-A22B Instruct 2507 up to 16,384, DeepSeek-V3 up to 8,192 tokens.
Which can read images, PDFs, audio or video?
Llama 4 Maverick 17B Instruct accepts text and images; Qwen3 235B-A22B Instruct 2507 accepts text; DeepSeek-V3 accepts text. Llama 4 Maverick 17B Instruct handles the widest range of inputs.
Are any of these open source?
Yes, all three publish their weights (Apache 2.0 and DeepSeek Model License), so you can self-host them.
Which is newer?
Qwen3 235B-A22B Instruct 2507 is the newest, released Jul 21, 2025. Llama 4 Maverick 17B Instruct came out Apr 5, 2025; DeepSeek-V3 came out Dec 26, 2024. Knowledge cutoff: Llama 4 Maverick 17B Instruct Aug 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.
Related comparisons
- Llama 4 Maverick 17B Instruct vs Magistral SmallMistral AI · ECI 133.2
- Llama 4 Maverick 17B Instruct vs GPT-4.1 miniOpenAI · ECI 135.0
- Llama 4 Maverick 17B Instruct vs Qwen3 8BAlibaba (Qwen) · ECI 136.2
- Qwen3 235B-A22B Instruct 2507 vs GPT OSS 120BOpenAI · ECI 140.0
- Qwen3 235B-A22B Instruct 2507 vs DeepSeek-V3.1DeepSeek · ECI 139.9
- Qwen3 235B-A22B Instruct 2507 vs Gemini 2.5 FlashGoogle · ECI 140.8
