Gecko Tests · ScorecardAs of 2026-11-17 · 74 models tested · 56 with 3+ tests ranked
Gecko Scorecard · Who Passes Our Tests?
All of BenchGecko's own tests on one page, in plain words. Not how smart a model is, but how it behaves.
In plain wordsEach column is one test. A is best, E is worst. The Gecko Score adds them up: the higher, the better the model behaves on our tests.
Top 5 on the Gecko Tests
#1
Qwen3.8 Max Prime
83
Gecko Score · 3 tests
Best at: fair price in other languages
#2
Qwen3.8 27B
80
Gecko Score · 4 tests
Best at: knows recent news
#3
Claude Opus 5.5
70
Gecko Score · 5 tests
Best at: draws the world from memory
#4
Claude Sonnet 5.5
69
Gecko Score · 5 tests
Best at: knows recent news
#5
Qwen3.8 Max (0902)
68
Gecko Score · 5 tests
Best at: fair price in other languages
| # | Model | Gecko Score | Who Are You Knows who made it · 71 models | World Map Draws the world from memory · 13 models | Censorship Index Answers normal questions · 47 models | Knowledge Horizon Knows recent news · 55 models | Tokenizer Tax Fair price in other languages · 67 models |
|---|---|---|---|---|---|---|---|
| 1 | Qwen3.8 Max Prime | 83 | BKnows who made it | not tested yet | not tested yet | AKnows news up to May 2026 | A35% more tokens outside English |
| 2 | Qwen3.8 27B | 80 | BKnows who made it | not tested yet | AAnswers almost everything | AKnows news up to Apr 2026 | B41% more tokens outside English |
| 3 | Claude Opus 5.5 | 70 | BKnows who made it | A97.9% of the map right | AAnswers almost everything | AKnows news up to May 2026 | E73% more tokens outside English |
| 4 | Claude Sonnet 5.5 | 69 | BKnows who made it | A97.5% of the map right | AAnswers almost everything | AKnows news up to May 2026 | E73% more tokens outside English |
| 5 | Qwen3.8 Max (0902) | 68 | BKnows who made it | D93.9% of the map right | CAnswers almost everything | AKnows news up to May 2026 | A35% more tokens outside English |
| 6 | MiniMax M3 | 67 | BKnows who made it | D92.8% of the map right | AAnswers almost everything | BKnows news up to Nov 2025 | A36% more tokens outside English |
| 7 | Gemini 4 Hydrogen | 67 | BKnows who made it | not tested yet | AAnswers almost everything | CKnows news up to Feb 2026 | not tested yet |
| 8 | Qwen3.8 Flash | 67 | BKnows who made it | not tested yet | not tested yet | CKnows news up to Oct 2025 | A35% more tokens outside English |
| 9 | Qwen4 Max | 67 | BKnows who made it | not tested yet | AAnswers almost everything | CKnows news up to Feb 2026 | not tested yet |
| 10 | GPT-6.1 Sol | 66 | BKnows who made it | A98.0% of the map right | CAnswers almost everything | BKnows news up to Mar 2026 | C56% more tokens outside English |
| 11 | Qwen3.8 2.4T A95B | 63 | BKnows who made it | not tested yet | DRefuses or dodges 5% | BKnows news up to Mar 2026 | A39% more tokens outside English |
| 12 | GPT-6 Sol | 61 | DKnows who made it | not tested yet | AAnswers almost everything | BKnows news up to Apr 2026 | C53% more tokens outside English |
| 13 | GPT-6 Luna | 60 | BKnows who made it | C96.2% of the map right | CAnswers almost everything | BKnows news up to Apr 2026 | C56% more tokens outside English |
| 14 | GPT-6 Luna Pro | 60 | BKnows who made it | not tested yet | CAnswers almost everything | BKnows news up to Apr 2026 | D57% more tokens outside English |
| 15 | Mistral Small 4 | 58 | DSometimes says it is OpenAI | not tested yet | CAnswers almost everything | AKnows news up to Mar 2026 | C53% more tokens outside English |
| 16 | GPT-6.1 Sol Pro | 57 | BKnows who made it | not tested yet | CAnswers almost everything | BKnows news up to Mar 2026 | D64% more tokens outside English |
| 17 | DeepSeek V4 Flash | 56 | BKnows who made it | not tested yet | DRefuses or dodges 5% | DKnows news up to Apr 2025 | A7% more tokens outside English |
| 18 | Gemini 3.6 Flash | 56 | BKnows who made it | not tested yet | CAnswers almost everything | DKnows news up to Jul 2025 | B44% more tokens outside English |
| 19 | GPT-6.1 Terra Pro | 56 | DKnows who made it | not tested yet | AAnswers almost everything | BKnows news up to Apr 2026 | C55% more tokens outside English |
| 20 | GPT-6.1 Terra | 55 | DKnows who made it | not tested yet | AAnswers almost everything | BKnows news up to Apr 2026 | C56% more tokens outside English |
| 21 | Gemini 4 Carbon | 55 | BKnows who made it | not tested yet | ERefuses or dodges 5% | not tested yet | A39% more tokens outside English |
| 22 | Gemini 3.8 Flash | 54 | BKnows who made it | B97.4% of the map right | CAnswers almost everything | EKnows news up to Mar 2025 | B44% more tokens outside English |
| 23 | MiniMax M2-her | 54 | DKnows who made it | not tested yet | ERefuses or dodges 8% | AKnows news up to Oct 2025 | A36% more tokens outside English |
| 24 | GLM 5.3 FlashX Legal | 54 | BKnows who made it | not tested yet | CAnswers almost everything | CKnows news up to Jan 2026 | not tested yet |
| 25 | Grok Build 0.1 | 54 | BKnows who made it | not tested yet | not tested yet | DKnows news up to Feb 2025 | B44% more tokens outside English |
| 26 | Grok 4.3 | 53 | BKnows who made it | not tested yet | not tested yet | DKnows news up to Dec 2024 | B44% more tokens outside English |
| 27 | Mistral Large 4 | 53 | BKnows who made it | not tested yet | AAnswers almost everything | EKnows news up to Apr 2025 | not tested yet |
| 28 | Claude Opus 5 | 52 | BKnows who made it | not tested yet | DAnswers almost everything | AKnows news up to Mar 2026 | E73% more tokens outside English |
| 29 | Gemini 3.5 Flash Lite | 51 | BKnows who made it | not tested yet | CAnswers almost everything | EKnows news up to Feb 2025 | B44% more tokens outside English |
| 30 | Gemini 3.7 Flash | 51 | BKnows who made it | not tested yet | CAnswers almost everything | EKnows news up to Mar 2025 | B44% more tokens outside English |
| 31 | MiniMax M3.1 Flash | 51 | ESometimes says it is OpenAI or Anthropic | not tested yet | DAnswers almost everything | BKnows news up to May 2026 | A36% more tokens outside English |
| 32 | Grok 4.7 | 50 | BKnows who made it | B97.3% of the map right | ERefuses or dodges 5% | DKnows news up to Oct 2025 | B50% more tokens outside English |
| 33 | Claude Haiku 5.5 | 49 | BKnows who made it | not tested yet | CAnswers almost everything | BKnows news up to Mar 2026 | E73% more tokens outside English |
| 34 | GLM 5.3 Prime | 49 | BKnows who made it | not tested yet | CAnswers almost everything | CKnows news up to Dec 2025 | D70% more tokens outside English |
| 35 | Kimi K2.7 Code | 48 | BKnows who made it | not tested yet | AAnswers almost everything | DKnows news up to Jun 2025 | E92% more tokens outside English |
| 36 | GLM 5.3 | 47 | BKnows who made it | not tested yet | CAnswers almost everything | CKnows news up to Dec 2025 | E90% more tokens outside English |
| 37 | Grok 4.5 | 46 | BKnows who made it | not tested yet | not tested yet | EKnows news up to Feb 2025 | B50% more tokens outside English |
| 38 | GLM 5.3 Flash | 45 | BKnows who made it | E89.7% of the map right | CAnswers almost everything | BKnows news up to Jan 2026 | D70% more tokens outside English |
| 39 | Devstral 2 2512 | 45 | ESometimes says it is OpenAI | not tested yet | AAnswers almost everything | DKnows news up to Nov 2024 | C53% more tokens outside English |
| 40 | MiniMax M2.7 | 44 | BKnows who made it | not tested yet | ERefuses or dodges 13% | EKnows news up to Oct 2024 | A35% more tokens outside English |
| 41 | Gemini 4 Argon | 43 | BKnows who made it | not tested yet | DRefuses or dodges 5% | EKnows news up to Feb 2025 | A39% more tokens outside English |
| 42 | Kimi K3.1 Pro | 43 | BKnows who made it | not tested yet | DAnswers almost everything | BKnows news up to Apr 2026 | E92% more tokens outside English |
| 43 | Grok 4.6 | 43 | BKnows who made it | not tested yet | not tested yet | EKnows news up to Jan 2025 | B50% more tokens outside English |
| 44 | Kimi K2.5 | 42 | BKnows who made it | not tested yet | CAnswers almost everything | CKnows news up to Apr 2025 | E100% more tokens outside English |
| 45 | Kimi K3 | 41 | BKnows who made it | C94.6% of the map right | not tested yet | CKnows news up to Nov 2025 | E93% more tokens outside English |
| 46 | DeepSeek V4 Pro | 38 | BKnows who made it | C95.8% of the map right | ERefuses or dodges 8% | DKnows news up to Apr 2025 | D68% more tokens outside English |
| 47 | DeepSeek V4 Pro 0813 | 38 | ESometimes says it is OpenAI or Anthropic | not tested yet | ERefuses or dodges 10% | CKnows news up to Nov 2025 | A17% more tokens outside English |
| 48 | GLM 5.3 FlashX | 37 | BKnows who made it | not tested yet | ERefuses or dodges 10% | CKnows news up to Jan 2026 | D70% more tokens outside English |
| 49 | DeepSeek V4.1 Flash | 36 | ESometimes says it is OpenAI | not tested yet | ERefuses or dodges 6% | AKnows news up to May 2026 | D68% more tokens outside English |
| 50 | Mistral Medium 3.5 | 34 | ESometimes says it is OpenAI | not tested yet | CAnswers almost everything | EKnows news up to Nov 2024 | C53% more tokens outside English |
| 51 | DeepSeek V4.1 Flash 1017 | 32 | ESometimes says it is OpenAI | not tested yet | ERefuses or dodges 5% | BKnows news up to May 2026 | D68% more tokens outside English |
| 52 | Kimi K2.6 | 28 | BKnows who made it | not tested yet | DRefuses or dodges 5% | EKnows news up to Nov 2024 | E92% more tokens outside English |
| 53 | Llama 4 Maverick | 25 | ESometimes says it is Google or OpenAI | E79.2% of the map right | not tested yet | not tested yet | B44% more tokens outside English |
| 54 | DeepSeek V4 Flash 0731 | 24 | ESometimes says it is Google or Anthropic | not tested yet | DRefuses or dodges 5% | DKnows news up to Jun 2025 | D68% more tokens outside English |
| 55 | GLM 5.2 | 23 | DSometimes says it is Google | not tested yet | not tested yet | DKnows news up to Feb 2025 | D70% more tokens outside English |
| 56 | MiniMax M2.5 | 21 | ESometimes says it is OpenAI | not tested yet | ERefuses or dodges 8% | DKnows news up to Dec 2024 | D64% more tokens outside English |
| Claude Fable 5.1 | 2/3 tests | BKnows who made it | not tested yet | not tested yet | not tested yet | E73% more tokens outside English | |
| Claude Fable 5.2 | 2/3 tests | BKnows who made it | not tested yet | not tested yet | not tested yet | E73% more tokens outside English | |
| Claude Opus 5 (Fast) | 2/3 tests | BKnows who made it | not tested yet | not tested yet | not tested yet | E73% more tokens outside English | |
| Gemma 4 26B A4B | 2/3 tests | BKnows who made it | not tested yet | not tested yet | not tested yet | B44% more tokens outside English | |
| GLM 5.1 | 2/3 tests | BKnows who made it | not tested yet | not tested yet | not tested yet | D71% more tokens outside English | |
| Grok 4.20 | 2/3 tests | BKnows who made it | not tested yet | not tested yet | not tested yet | B44% more tokens outside English | |
| Kimi K2 0905 | 2/3 tests | BKnows who made it | not tested yet | not tested yet | not tested yet | E92% more tokens outside English | |
| Kimi K2 Thinking | 2/3 tests | BKnows who made it | not tested yet | not tested yet | not tested yet | E92% more tokens outside English | |
| Llama 4 Scout | 2/3 tests | DSometimes says it is OpenAI | not tested yet | not tested yet | not tested yet | B43% more tokens outside English | |
| MiniMax M2 | 2/3 tests | ESometimes says it is Anthropic | not tested yet | not tested yet | not tested yet | A36% more tokens outside English | |
| MiniMax M2.1 | 2/3 tests | ESometimes says it is Anthropic | not tested yet | not tested yet | not tested yet | A36% more tokens outside English | |
| Ministral 3 14B 2512 | 2/3 tests | DSometimes says it is OpenAI | not tested yet | not tested yet | not tested yet | C53% more tokens outside English | |
| Ministral 3 3B 2512 | 2/3 tests | ESometimes says it is OpenAI | not tested yet | not tested yet | not tested yet | C53% more tokens outside English | |
| Nano Banana 2.1 | 2/3 tests | BKnows who made it | not tested yet | not tested yet | EKnows news up to Feb 2025 | not tested yet | |
| Gemini 3.5 Flash | 1/3 tests | BKnows who made it | not tested yet | not tested yet | not tested yet | not tested yet | |
| GPT-6.2 Sol | 1/3 tests | not tested yet | not tested yet | not tested yet | not tested yet | C56% more tokens outside English | |
| GPT-6.2 Sol Pro | 1/3 tests | not tested yet | not tested yet | not tested yet | not tested yet | C54% more tokens outside English | |
| Mistral Large | 1/3 tests | not tested yet | E77.5% of the map right | not tested yet | not tested yet | not tested yet |
Every new model from a tracked lab is tested automatically within a day of release; grades update as models are added. Method per test on each test page.
How to cite · data as of 2026-11-17
BenchGecko Gecko Scorecard. BenchGecko, data as of 2026-11-17. https://benchgecko.ai/gecko-tests/scorecard
BenchGecko measures this data itself: free to reuse under CC BY 4.0 with the credit "Source: BenchGecko" and a link. JSON · llms.txt · MCP
Frequently Asked Questions
The average of a model's rank across BenchGecko's own tests (0 to 100, higher is better). Each test is ranked separately, so a model is compared with the other models that took the same test. A model needs at least 3 tests to get a score.