Gemma 3 27B IT vs GPT-4.1 nano vs Qwen3.5 9B
Qwen3.5 9B comes out ahead, 72 to 63 and 60 on our weighted score.
Google
Gemma 3 27B IT
60/100- ECI130.0
- Price$0.08 / $0.20
- Context131K
OpenAI
GPT-4.1 nano
63/100- ECI129.6
- Price$0.10 / $0.40
- Context1.05M
- Our pick
Alibaba (Qwen)
Qwen3.5 9B
72/100- ECI139.5
- Price$0.10 / $0.15
- Context262K
Qwen3.5 9B is our pick
Qwen3.5 9B is the better all-round choice, scoring 72/100 against GPT-4.1 nano (63) and Gemma 3 27B IT (60). It leads on capability and inputs & features. GPT-4.1 nano wins on context window. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
- CapabilityQwen3.5 9BCapabilities Index (ECI): Qwen3.5 9B 139.5 · Gemma 3 27B IT 130.0 · GPT-4.1 nano 129.6
- Lowest priceGemma 3 27B ITGemma 3 27B IT $0.11 · Qwen3.5 9B $0.113 · GPT-4.1 nano $0.175 per 1M tokens (3:1 blend)
- Longest contextGPT-4.1 nanoGPT-4.1 nano 1,047,576 · Qwen3.5 9B 262,144 · Gemma 3 27B IT 131,072 tokens
- Widest inputsQwen3.5 9BGemma 3 27B IT: Text, Images · GPT-4.1 nano: Text, Images · Qwen3.5 9B: Text, Images, Video
- Self-hostingGemma 3 27B IT and Qwen3.5 9BPublishes downloadable weights
| Measure | Weight | Gemma 3 27B IT | GPT-4.1 nano | Qwen3.5 9B |
|---|---|---|---|---|
| CapabilityCapabilities Index (ECI) | 50% | 53 | 52 | 65 |
| Price | 25% | 95 | 86 | 95 |
| Inputs & features | 15% | 50 | 60 | 80 |
| Context window | 10% | 24 | 61 | 37 |
| Overall | 100% | 60/100 | 63/100 | 72/100 |
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | |||
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 130.0 | 129.6 | 139.5 (best) |
| ECI rank | #125 of 148 | #127 of 148 | #101 of 148 (best) |
| GPQA DiamondGraduate-level science questions | 47.7% | 48.9% | 79.0% (best) |
| OTIS Mock AIME 2024–2025Competition mathematics | 22.5% | 28.9% | 61.7% (best) |
| SimpleQA VerifiedShort factual questions | — | 6.0% | — |
| Price per million tokens | |||
| Input | $0.08 (best) | $0.10 | $0.10 |
| Output | $0.20 | $0.40 | $0.15 (best) |
| Cached input | — | $0.025 | — |
| Blended (3:1) | $0.11 (best) | $0.175 | $0.113 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Median of 9 providers | Official OpenAI API | Median of 14 providers |
| Limits | |||
| Context window | 131,072 tokens | 1,047,576 tokens (best) | 262,144 tokens |
| Max output | 131,072 tokens (best) | 32,768 tokens | 65,536 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | Yes | Yes | Yes |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | No | Yes |
| Reasoning | No | No | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | Yes | Yes |
| Availability | |||
| Weights | Open | Proprietary | Open |
| API model ID | — | gpt-4.1-nano | — |
| API providers | 10 | 20 (best) | 15 |
| Released | Mar 12, 2025 | Apr 14, 2025 | Feb 23, 2026 |
| Knowledge cutoff | Aug 2024 | Apr 2024 | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
Gemma 3 27B IT$1.20
GPT-4.1 nano$1.80
Qwen3.5 9B$1.30
Which should you choose?
Which is better: Gemma 3 27B IT, GPT-4.1 nano or Qwen3.5 9B?
Qwen3.5 9B is the better all-round choice, scoring 72/100 against GPT-4.1 nano (63) and Gemma 3 27B IT (60). It leads on capability and inputs & features. GPT-4.1 nano wins on context window. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
Which is cheaper, Gemma 3 27B IT, GPT-4.1 nano or Qwen3.5 9B?
Gemma 3 27B IT is cheaper at $0.08 input / $0.20 output per million tokens (median across 9 API providers). Qwen3.5 9B costs $0.10 input / $0.15 output per million tokens (median across 14 API providers); GPT-4.1 nano costs $0.10 input / $0.40 output per million tokens (official OpenAI API price). At a typical mix of three input tokens to one output token, that is $0.11 per million tokens for Gemma 3 27B IT versus $0.113 for Qwen3.5 9B (1× as much) and $0.175 for GPT-4.1 nano (1.6× as much).
Which scores higher on benchmarks?
Qwen3.5 9B scores higher on the Capabilities Index (ECI): Qwen3.5 9B 139.5 (#101 of 148), Gemma 3 27B IT 130.0 (#125 of 148) and GPT-4.1 nano 129.6 (#127 of 148). Their confidence ranges do not overlap (136.5–141.3 vs 124.1–132.2), so the gap is a real one. On individual benchmarks: GPQA Diamond — Qwen3.5 9B 79.0%, GPT-4.1 nano 48.9%, Gemma 3 27B IT 47.7%; OTIS Mock AIME 2024–2025 — Qwen3.5 9B 61.7%, GPT-4.1 nano 28.9%, Gemma 3 27B IT 22.5%.
Which is better for coding?
There are no published SWE-bench Verified results for Gemma 3 27B IT, GPT-4.1 nano and Qwen3.5 9B yet, so there is no like-for-like coding score. On overall capability, Qwen3.5 9B 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?
GPT-4.1 nano has the largest context window at 1,047,576 tokens, against 262,144 for Qwen3.5 9B and 131,072 for Gemma 3 27B IT. Maximum output per response: Gemma 3 27B IT up to 131,072, GPT-4.1 nano up to 32,768, Qwen3.5 9B up to 65,536 tokens.
Which can read images, PDFs, audio or video?
Gemma 3 27B IT accepts text and images; GPT-4.1 nano accepts text and images; Qwen3.5 9B accepts text, images and video. Qwen3.5 9B handles the widest range of inputs.
Are any of these open source?
Gemma 3 27B IT and Qwen3.5 9B publishes its weights and can be self-hosted; GPT-4.1 nano is proprietary.
Which is newer?
Qwen3.5 9B is the newest, released Feb 23, 2026. GPT-4.1 nano came out Apr 14, 2025; Gemma 3 27B IT came out Mar 12, 2025. Knowledge cutoff: Gemma 3 27B IT Aug 2024, GPT-4.1 nano Apr 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.