Qwen Turbo vs Gemma 3 12B IT vs Claude Haiku 3.5
Too close to call on our weighted score (Gemma 3 12B IT 32, Qwen Turbo 31, Claude Haiku 3.5 30). The right pick depends on what you value most.
Alibaba (Qwen)Qwen Turbo
31/100- ECI—
- Price$0.05 / $0.20
- Context1M
GoogleGemma 3 12B IT
32/100- ECI123.5
- Price$0.05 / $0.15
- Context131K
AnthropicClaude Haiku 3.5
30/100- ECI127.2
- Price—
- Context200K
Too close to call
It is close. Our weighted score puts them within 1 points (Gemma 3 12B IT 32/100, Qwen Turbo 31/100, Claude Haiku 3.5 30/100), so choose by what matters most for your work: Gemma 3 12B IT for raw capability and Qwen Turbo for long inputs. The score weighs capability 67%, inputs & features 20%, context window 13%. Capability uses the average of shared benchmarks (GPQA Diamond and OTIS Mock AIME 2024–2025), because Qwen Turbo has no Capabilities Index score yet.
- CapabilityGemma 3 12B ITShared benchmarks: Gemma 3 12B IT 28.1% · Qwen Turbo 24.0% · Claude Haiku 3.5 21.2%
- Lowest priceGemma 3 12B ITGemma 3 12B IT $0.075 · Qwen Turbo $0.087 per 1M tokens (3:1 blend) · Claude Haiku 3.5 unpriced
- Longest contextQwen TurboQwen Turbo 1,000,000 · Claude Haiku 3.5 200,000 · Gemma 3 12B IT 131,072 tokens
- Widest inputsClaude Haiku 3.5Qwen Turbo: Text · Gemma 3 12B IT: Text, Images · Claude Haiku 3.5: Text, Images, PDFs
- Self-hostingGemma 3 12B ITPublishes downloadable weights
| Measure | Weight | Qwen Turbo | Gemma 3 12B IT | Claude Haiku 3.5 |
|---|---|---|---|---|
| CapabilityShared benchmarks | 67% | 24 | 28 | 21 |
| Inputs & features | 20% | 35 | 50 | 60 |
| Context window | 13% | 60 | 24 | 32 |
| Overall | 100% | 31/100 | 32/100 | 30/100 |
Left out because at least one model lacks the data: price. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | Qwen TurboAlibaba (Qwen) | Gemma 3 12B ITGoogle | Claude Haiku 3.5Anthropic |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | 123.5 | 127.2 (best) |
| ECI rank | — | #138 of 148 | #134 of 148 (best) |
| GPQA DiamondGraduate-level science questions | 41.8% (best) | 39.5% | 38.1% |
| OTIS Mock AIME 2024–2025Competition mathematics | 6.1% | 16.7% (best) | 4.3% |
| Price per million tokens | |||
| Input | $0.05 | $0.05 | — |
| Output | $0.20 | $0.15 (best) | — |
| Cached input | — | — | — |
| Blended (3:1) | $0.087 | $0.075 (best) | — |
| Long-context rate | Same rate | Same rate | — |
| Price source | Official Alibaba API | Median of 7 providers | — |
| Limits | |||
| Context window | 1,000,000 tokens (best) | 131,072 tokens | 200,000 tokens |
| Max output | 16,384 tokens | 131,072 tokens (best) | 8,192 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | Yes |
| PDFs | No | No | Yes |
| Audio | No | No | No |
| Video | No | No | No |
| Reasoning | Yes | No | No |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | No | No |
| Availability | |||
| Weights | Proprietary | Open | Proprietary |
| API model ID | qwen-turbo | — | — |
| API providers | 3 | 7 (best) | — |
| Released | Nov 1, 2024 | Mar 12, 2025 | Oct 22, 2024 |
| Knowledge cutoff | Apr 2024 | Aug 2024 | Jul 31, 2024 |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- Qwen Turbo$0.90
- Gemma 3 12B IT$0.80
- Claude Haiku 3.5—
Which should you choose?
Which is better: Qwen Turbo, Gemma 3 12B IT or Claude Haiku 3.5?
It is close. Our weighted score puts them within 1 points (Gemma 3 12B IT 32/100, Qwen Turbo 31/100, Claude Haiku 3.5 30/100), so choose by what matters most for your work: Gemma 3 12B IT for raw capability and Qwen Turbo for long inputs. The score weighs capability 67%, inputs & features 20%, context window 13%. Capability uses the average of shared benchmarks (GPQA Diamond and OTIS Mock AIME 2024–2025), because Qwen Turbo has no Capabilities Index score yet.
Which is cheaper, Qwen Turbo, Gemma 3 12B IT or Claude Haiku 3.5?
Gemma 3 12B IT is cheaper at $0.05 input / $0.15 output per million tokens (median across 7 API providers). Qwen Turbo costs $0.05 input / $0.20 output per million tokens (official Alibaba API price). At a typical mix of three input tokens to one output token, that is $0.075 per million tokens for Gemma 3 12B IT versus $0.087 for Qwen Turbo (1.2× as much). Claude Haiku 3.5 has no published per-token price.
Which scores higher on benchmarks?
Not every model here has a Capabilities Index score, so we compare the average of shared benchmarks (GPQA Diamond and OTIS Mock AIME 2024–2025): Gemma 3 12B IT 28.1%, Qwen Turbo 24.0% and Claude Haiku 3.5 21.2%. On individual benchmarks: GPQA Diamond — Qwen Turbo 41.8%, Gemma 3 12B IT 39.5%, Claude Haiku 3.5 38.1%; OTIS Mock AIME 2024–2025 — Gemma 3 12B IT 16.7%, Qwen Turbo 6.1%, Claude Haiku 3.5 4.3%.
Which is better for coding?
There are no published SWE-bench Verified results for Qwen Turbo, Gemma 3 12B IT and Claude Haiku 3.5 yet, so there is no like-for-like coding score. On overall capability, Gemma 3 12B IT 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?
Qwen Turbo has the largest context window at 1,000,000 tokens, against 200,000 for Claude Haiku 3.5 and 131,072 for Gemma 3 12B IT. Maximum output per response: Qwen Turbo up to 16,384, Gemma 3 12B IT up to 131,072, Claude Haiku 3.5 up to 8,192 tokens.
Which can read images, PDFs, audio or video?
Qwen Turbo accepts text; Gemma 3 12B IT accepts text and images; Claude Haiku 3.5 accepts text, images and PDFs. Claude Haiku 3.5 handles the widest range of inputs.
Are any of these open source?
Gemma 3 12B IT publishes its weights and can be self-hosted; Qwen Turbo and Claude Haiku 3.5 is proprietary.
Which is newer?
Gemma 3 12B IT is the newest, released Mar 12, 2025. Qwen Turbo came out Nov 1, 2024; Claude Haiku 3.5 came out Oct 22, 2024. Knowledge cutoff: Qwen Turbo Apr 2024, Gemma 3 12B IT Aug 2024, Claude Haiku 3.5 Jul 31, 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.
