# Lab score: open-weight models on our own evaluation (slim)

> Open-weight models answering 59 questions about Anchor Terminal through its MCP server, pages and API, scored 0 to 100, with every run dated.

- Full: https://www.anchorterminal.com/benchmark/lab/evaluation/models/index.md (~1,600 tokens) · this version ~1,430 tokens · JSON https://www.anchorterminal.com/benchmark/lab/evaluation/models/index.json · canonical https://www.anchorterminal.com/benchmark/lab/evaluation/models/
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · MCP: https://www.anchorterminal.com/mcp · Updated: 2026-11-22
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

Each open-weight model the lab runs answers 59 questions about Anchor Terminal through our MCP server, pages and API, three times over, calling the tools itself or writing programs that call them, as it chooses.

## Each model's latest

| # | Model | Lab score | Right | One lookup | A few | Hard | How it got there | A question takes | Tokens a second | Date |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | [Qwen3.8-Flash-Next](https://www.anchorterminal.com/tools/qwen3-8-flash-next.md) (qwen3.8-flash-next) | 98 | 57 of 59 | 18 of 20 | 27 of 27 | 12 of 12 | own calls 56%, both 44% | 71 s | 18.6 | 20 November 2026 |
| 2 | [Gemma 4 26B A4B IT](https://www.anchorterminal.com/tools/gemma-4-26b-a4b-it.md) (gemma-4-26b-a4b) | 92 | 55 of 59 | 19 of 20 | 26 of 27 | 10 of 12 | own calls 57%, programs 1%, both 43% | 91 s | 46.5 | 19 November 2026 |
| 3 | [Gemma 4 31B IT](https://www.anchorterminal.com/tools/gemma-4-31b-it.md) (gemma-4-31b) | 87 | 52 of 59 | 18 of 20 | 24 of 27 | 10 of 12 | own calls 56%, both 44% | 5.3 min | 9.1 | 19 November 2026 |
| 4 | [Qwen3.6-35B-A3B](https://www.anchorterminal.com/tools/qwen3-6-35b-a3b.md) (qwen3.6-35b-a3b) | 77 | 50 of 59 | 20 of 20 | 25 of 27 | 5 of 12 | own calls 75%, both 25% | 56 s | 33.8 | 20 November 2026 |
| 5 | [GLM-4.5-Air](https://www.anchorterminal.com/tools/glm-4-5-air.md) (glm-4.5-air) | 67 | 44 of 59 | 19 of 20 | 20 of 27 | 5 of 12 | own calls 82%, both 18% | 2.3 min | 9.4 | 21 November 2026 |
| 6 | [gpt-oss-120b](https://www.anchorterminal.com/tools/gpt-oss-120b.md) (gpt-oss-120b) | 66 | 43 of 59 | 18 of 20 | 20 of 27 | 5 of 12 | own calls 71%, both 29% | 55 s | 29.9 | 20 November 2026 |
| 7 | [Qwen3.5-9B](https://www.anchorterminal.com/tools/qwen3-5-9b.md) (qwen3.5-9b) | 66 | 44 of 59 | 18 of 20 | 23 of 27 | 3 of 12 | own calls 72%, both 28% | 2.0 min | 17.2 | 21 November 2026 |
| 8 | [Nemotron 3 Super 120B A12B](https://www.anchorterminal.com/tools/nemotron-3-super-120b-a12b.md) (nemotron-3-super-120b-a12b) | 63 | 41 of 59 | 18 of 20 | 18 of 27 | 5 of 12 | own calls 60%, both 40% | 4.8 min | 15.0 | 21 November 2026 |
| 9 | [Qwen3.5-4B](https://www.anchorterminal.com/tools/qwen3-5-4b.md) (qwen3.5-4b) | 62 | 41 of 59 | 18 of 20 | 19 of 27 | 4 of 12 | own calls 83%, both 17% | 69 s | 26.3 | 21 November 2026 |
| 10 | [gpt-oss-20b](https://www.anchorterminal.com/tools/gpt-oss-20b.md) (gpt-oss-20b) | 61 | 40 of 59 | 17 of 20 | 19 of 27 | 4 of 12 | own calls 69%, both 31% | 48 s | 45.0 | 21 November 2026 |
| 11 | [Qwen3-Coder-Next](https://www.anchorterminal.com/tools/qwen3-coder-next.md) (qwen3-coder-next) | 61 | 41 of 59 | 18 of 20 | 20 of 27 | 3 of 12 | own calls 54%, both 46% | 53 s | 32.7 | 22 November 2026 |
| 12 | [Qwen3-Coder-30B-A3B-Instruct](https://www.anchorterminal.com/tools/qwen3-coder-30b-a3b-instruct.md) (qwen3-coder-30b-a3b) | 59 | 38 of 59 | 14 of 20 | 21 of 27 | 3 of 12 | own calls 66%, both 34% | 64 s | 34.9 | 22 November 2026 |
| 13 | [Nemotron 3 Nano 30B A3B](https://www.anchorterminal.com/tools/nemotron-3-nano-30b-a3b.md) (nemotron-3-nano-30b-a3b) | 59 | 40 of 59 | 18 of 20 | 19 of 27 | 3 of 12 | own calls 70%, both 30% | 3.2 min | 48.5 | 22 November 2026 |
| 14 | [GLM-4.7-Flash](https://www.anchorterminal.com/tools/glm-4-7-flash.md) (glm-4.7-flash) | 55 | 38 of 59 | 19 of 20 | 16 of 27 | 3 of 12 | own calls 75%, programs 1%, both 25% | 102 s | 14.4 | 20 November 2026 |
| 15 | [Qwen3-30B-A3B-Instruct-2507](https://www.anchorterminal.com/tools/qwen3-30b-a3b-instruct-2507.md) (qwen3-30b-a3b-instruct-2507) | 47 | 31 of 59 | 14 of 20 | 13 of 27 | 4 of 12 | own calls 49%, both 51% | 68 s | 31.7 | 22 November 2026 |

## Every evaluation

A run in code mode, where the model may only write programs, is kept beside the Lab score's runs and scores nothing.

| Date | Model | Mode | Score | Right |
| --- | --- | --- | --- | --- |
| 22 November 2026 | qwen3-coder-30b-a3b | either | 59 | 38 of 59 |
| 22 November 2026 | qwen3-30b-a3b-instruct-2507 | either | 47 | 31 of 59 |
| 22 November 2026 | qwen3-coder-next | either | 61 | 41 of 59 |
| 22 November 2026 | nemotron-3-nano-30b-a3b | either | 59 | 40 of 59 |
| 21 November 2026 | glm-4.5-air | either | 67 | 44 of 59 |
| 21 November 2026 | gpt-oss-20b | either | 61 | 40 of 59 |
| 21 November 2026 | qwen3.5-4b | either | 62 | 41 of 59 |
| 21 November 2026 | qwen3.5-9b | either | 66 | 44 of 59 |
| 21 November 2026 | nemotron-3-super-120b-a12b | either | 63 | 41 of 59 |
| 20 November 2026 | qwen3.8-flash-next | either | 98 | 57 of 59 |
| 20 November 2026 | gpt-oss-120b | either | 66 | 43 of 59 |
| 20 November 2026 | qwen3.6-35b-a3b | either | 77 | 50 of 59 |
| 20 November 2026 | glm-4.7-flash | either | 55 | 38 of 59 |
| 19 November 2026 | gemma-4-26b-a4b | either | 92 | 55 of 59 |
| 19 November 2026 | gemma-4-31b | either | 87 | 52 of 59 |
| 18 November 2026 | qwen3.8-27b | code only | 89 | 54 of 59 |

The harness is open source (Go, standard library only, MIT): https://github.com/Anchor-Terminal/anchorlab

The questions: https://github.com/Anchor-Terminal/anchorlab/blob/main/eval/tasks.go

An answer is checked against the site before and after the model answered, a question whose subject left the site or that has two right answers is skipped, and an answer with no lookup behind it counts as wrong.

JSON: https://www.anchorterminal.com/api/v1/lab-evals.json
