Kolibri 1 vs Nemotron 3.5 Super VL 120B A12B vs Qwen3.8 Flash
Qwen3.8 Flash comes out ahead, 74 to 71 and 60 on our weighted score, and it is the cheaper option too.
Aleph AlphaKolibri 1
60/100- ECI—
- Price$0.164 / $0.574
- Context1.05M
NVIDIANemotron 3.5 Super VL 120B A12B
71/100- ECI—
- Price$0.15 / $0.50
- Context1.05M
- Our pick
Alibaba (Qwen)Qwen3.8 Flash
74/100- ECI—
- Price$0.15 / $0.47
- Context1M
Qwen3.8 Flash is our pick
Qwen3.8 Flash is the better all-round choice, scoring 74/100 against Nemotron 3.5 Super VL 120B A12B (71) and Kolibri 1 (60). It leads on inputs & features. 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 priceQwen3.8 FlashQwen3.8 Flash $0.23 · Nemotron 3.5 Super VL 120B A12B $0.237 · Kolibri 1 $0.266 per 1M tokens (3:1 blend)
- Longest contextKolibri 1 and Nemotron 3.5 Super VL 120B A12BKolibri 1 1,048,576 · Nemotron 3.5 Super VL 120B A12B 1,048,576 · Qwen3.8 Flash 1,000,000 tokens
- Widest inputsNemotron 3.5 Super VL 120B A12B and Qwen3.8 FlashKolibri 1: Text · Nemotron 3.5 Super VL 120B A12B: Text, Images, Video · Qwen3.8 Flash: Text, Images, Video
- Self-hostingKolibri 1 and Nemotron 3.5 Super VL 120B A12BPublishes downloadable weights (Apache-2.0)
| Measure | Weight | Kolibri 1 | Nemotron 3.5 Super VL 120B A12B | Qwen3.8 Flash |
|---|---|---|---|---|
| Price | 50% | 74 | 76 | 77 |
| Inputs & features | 30% | 35 | 70 | 80 |
| Context window | 20% | 61 | 61 | 60 |
| Overall | 100% | 60/100 | 71/100 | 74/100 |
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 | Kolibri 1Aleph Alpha | Nemotron 3.5 Super VL 120B A12BNVIDIA | Qwen3.8 FlashAlibaba (Qwen) |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $0.164 | $0.15 (best) | $0.15 (best) |
| Output | $0.574 | $0.50 | $0.47 (best) |
| Cached input | — | — | $0.016 |
| Blended (3:1) | $0.266 | $0.237 | $0.23 (best) |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Median of 4 providers | Median of 11 providers | Official Alibaba API |
| Limits | |||
| Context window | 1,048,576 tokens (best) | 1,048,576 tokens (best) | 1,000,000 tokens |
| Max output | — | 8,192 tokens | 131,072 tokens (best) |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | Yes |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | Yes | Yes |
| Reasoning | Yes | Yes | Yeslow · medium · xhigh |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | No | Yes |
| Availability | |||
| Weights | OpenApache-2.0 | Open | Proprietary |
| API model ID | — | nvidia/nemotron-3.5-super-vl-120b-a12b | qwen3.8-flash |
| API providers | 4 | 13 | 28 (best) |
| Released | Oct 3, 2026 | Oct 15, 2026 | Aug 26, 2026 |
| Knowledge cutoff | Jun 18, 2026 | — | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- Kolibri 1$2.78
- Nemotron 3.5 Super VL 120B A12B$2.50
- Qwen3.8 Flash$2.44
Which should you choose?
Which is better: Kolibri 1, Nemotron 3.5 Super VL 120B A12B or Qwen3.8 Flash?
Qwen3.8 Flash is the better all-round choice, scoring 74/100 against Nemotron 3.5 Super VL 120B A12B (71) and Kolibri 1 (60). It leads on inputs & features. 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, Kolibri 1, Nemotron 3.5 Super VL 120B A12B or Qwen3.8 Flash?
Qwen3.8 Flash is cheaper at $0.15 input / $0.47 output per million tokens (official Alibaba API price). Nemotron 3.5 Super VL 120B A12B costs $0.15 input / $0.50 output per million tokens (median across 11 API providers; free on Nvidia); Kolibri 1 costs $0.164 input / $0.574 output per million tokens (median across 4 API providers). At a typical mix of three input tokens to one output token, that is $0.23 per million tokens for Qwen3.8 Flash versus $0.237 for Nemotron 3.5 Super VL 120B A12B (1× as much) and $0.266 for Kolibri 1 (1.2× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Kolibri 1 has not been scored yet, Nemotron 3.5 Super VL 120B A12B has not been scored yet and Qwen3.8 Flash has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for Kolibri 1, Nemotron 3.5 Super VL 120B A12B and Qwen3.8 Flash 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?
Kolibri 1 and Nemotron 3.5 Super VL 120B A12B have the largest context windows (1,048,576 and 1,048,576 tokens), against 1,000,000 for Qwen3.8 Flash. Maximum output per response: Nemotron 3.5 Super VL 120B A12B up to 8,192, Qwen3.8 Flash up to 131,072 tokens.
Which can read images, PDFs, audio or video?
Kolibri 1 accepts text; Nemotron 3.5 Super VL 120B A12B accepts text, images and video; Qwen3.8 Flash accepts text, images and video. Nemotron 3.5 Super VL 120B A12B handles the widest range of inputs.
Are any of these open source?
Kolibri 1 and Nemotron 3.5 Super VL 120B A12B publishes its weights (Apache-2.0) and can be self-hosted; Qwen3.8 Flash is proprietary.
Which is newer?
Nemotron 3.5 Super VL 120B A12B is the newest, released Oct 15, 2026. Kolibri 1 came out Oct 3, 2026; Qwen3.8 Flash came out Aug 26, 2026. Knowledge cutoff: Kolibri 1 Jun 18, 2026.
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.
