Inkling vs Inkling Small vs Qwen3.8 27B
Inkling Small comes out ahead, 70 to 67 and 59 on our weighted score, and it is the cheaper option too.
Thinking MachinesInkling
59/100- ECI148.6
- Price$3.74 / $9.36
- Context1.05M
- Our pickThinking Machines
Inkling Small
70/100- ECI150.2
- Price$0.50 / $1.20
- Context1.05M
Alibaba (Qwen)Qwen3.8 27B
67/100- ECI149.4
- Price$0.40 / $2.50
- Context262K
Inkling Small is our pick
Inkling Small is the better all-round choice, scoring 70/100 against Qwen3.8 27B (67) and Inkling (59). It leads on price. Qwen3.8 27B wins on inputs & features. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
- CapabilityInkling SmallCapabilities Index (ECI): Inkling Small 150.2 · Qwen3.8 27B 149.4 · Inkling 148.6
- Lowest priceInkling SmallInkling Small $0.675 · Qwen3.8 27B $0.925 · Inkling $5.14 per 1M tokens (3:1 blend)
- Longest contextInkling and Inkling SmallInkling 1,048,576 · Inkling Small 1,048,576 · Qwen3.8 27B 262,144 tokens
- Widest inputsSame inputsInkling: Text, Images, Audio · Inkling Small: Text, Images, Audio · Qwen3.8 27B: Text, Images, Video
- Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
| Measure | Weight | Inkling | Inkling Small | Qwen3.8 27B |
|---|---|---|---|---|
| CapabilityCapabilities Index (ECI) | 50% | 76 | 78 | 77 |
| Price | 25% | 16 | 58 | 51 |
| Inputs & features | 15% | 70 | 70 | 80 |
| Context window | 10% | 61 | 61 | 37 |
| Overall | 100% | 59/100 | 70/100 | 67/100 |
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | InklingThinking Machines | Inkling SmallThinking Machines | Qwen3.8 27BAlibaba (Qwen) |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 148.6 | 150.2 (best) | 149.4 |
| ECI rank | #57 of 148 | #47 of 148 (best) | #53 of 148 |
| GPQA DiamondGraduate-level science questions | 88.3% | 88.5% (best) | — |
| FrontierMath Tiers 1–3Research-level mathematics | 33.3% | 46.3% (best) | — |
| OTIS Mock AIME 2024–2025Competition mathematics | 88.9% | 90.0% (best) | — |
| SimpleQA VerifiedShort factual questions | 40.3% (best) | 19.1% | — |
| Price per million tokens | |||
| Input | $3.74 | $0.50 | $0.40 (best) |
| Output | $9.36 | $1.20 (best) | $2.50 |
| Cached input | $0.748 | — | — |
| Blended (3:1) | $5.14 | $0.675 (best) | $0.925 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Official Thinking Machines API | Median of 11 providers | Median of 39 providers |
| Limits | |||
| Context window | 1,048,576 tokens (best) | 1,048,576 tokens (best) | 262,144 tokens |
| Max output | 1,048,576 tokens (best) | 1,048,576 tokens (best) | 32,768 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | Yes | Yes | Yes |
| PDFs | No | No | No |
| Audio | Yes | Yes | No |
| Video | No | No | Yes |
| Reasoning | Yeslow · medium · high · xhigh · max | Yes | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | No | Yes |
| Availability | |||
| Weights | OpenApache-2.0 | OpenApache-2.0 | Open |
| API model ID | thinkingmachines/Inkling:peft:262144 | — | — |
| API providers | 23 | 11 | 41 (best) |
| Released | Jul 15, 2026 | Jul 30, 2026 | Aug 14, 2026 |
| Knowledge cutoff | — | — | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- Inkling$56.12
- Inkling Small$7.40
- Qwen3.8 27B$9.00
Which should you choose?
Which is better: Inkling, Inkling Small or Qwen3.8 27B?
Inkling Small is the better all-round choice, scoring 70/100 against Qwen3.8 27B (67) and Inkling (59). It leads on price. Qwen3.8 27B wins on inputs & features. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
Which is cheaper, Inkling, Inkling Small or Qwen3.8 27B?
Inkling Small is cheaper at $0.50 input / $1.20 output per million tokens (median across 11 API providers). Qwen3.8 27B costs $0.40 input / $2.50 output per million tokens (median across 39 API providers); Inkling costs $3.74 input / $9.36 output per million tokens (official Thinking Machines API price). At a typical mix of three input tokens to one output token, that is $0.675 per million tokens for Inkling Small versus $0.925 for Qwen3.8 27B (1.4× as much) and $5.14 for Inkling (7.6× as much).
Which scores higher on benchmarks?
Inkling Small scores higher on the Capabilities Index (ECI): Inkling Small 150.2 (#47 of 148), Qwen3.8 27B 149.4 (#53 of 148) and Inkling 148.6 (#57 of 148). The confidence ranges of the top two overlap (147.5–152.1 vs 147.5–151.6), so treat the gap as small.
Which is better for coding?
There are no published SWE-bench Verified results for Inkling, Inkling Small and Qwen3.8 27B yet, so there is no like-for-like coding score. On overall capability, Inkling Small 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?
Inkling and Inkling Small have the largest context windows (1,048,576 and 1,048,576 tokens), against 262,144 for Qwen3.8 27B. Maximum output per response: Inkling up to 1,048,576, Inkling Small up to 1,048,576, Qwen3.8 27B up to 32,768 tokens.
Which can read images, PDFs, audio or video?
Inkling accepts text, images and audio; Inkling Small accepts text, images and audio; Qwen3.8 27B accepts text, images and video. They handle the same number of input types.
Are any of these open source?
Yes, all three publish their weights (Apache-2.0), so you can self-host them.
Which is newer?
Qwen3.8 27B is the newest, released Aug 14, 2026. Inkling Small came out Jul 30, 2026; Inkling came out Jul 15, 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 Oct 4, 2026.
