Сводный рейтинг Neira объединяет независимые тесты, сохраняя точные версии моделей и покрытие категорий.
Последний полный снимок: 23 авг. 2026 г., 06:02. Чат 30%, знания и рассуждение 20%, код 40%, вызов инструментов 10%.
Оценка 0–100 нормализована по позиции в каждом источнике, поэтому исходные шкалы не смешиваются.
| Ранг | Модель | Разработчик | Оценка | Покрытие | Источники |
|---|---|---|---|---|---|
#1 | Claude-Opus-4-5-20251101 (FC) | Anthropic | 100.0 | 1/1 | Berkeley Function Calling Leaderboard |
#2 | gpt-4.1-2025-04-14 | OpenAI | 100.0 | 1/1 | Galileo Agent Leaderboard v2 |
Открытые веса · 4495 моделей
#3 |
| Claude-Sonnet-4-5-20250929 (FC) |
| Anthropic |
| 99.1 |
| 1/1 |
| Berkeley Function Calling Leaderboard |
#4 | Gemini-3-Pro-Preview (Prompt) | 98.1 | 1/1 | Berkeley Function Calling Leaderboard |
#5 | GLM-4.6 (FC thinking) | zhipu-ai | 97.2 | 1/1 | Berkeley Function Calling Leaderboard |
#6 | Grok-4-1-fast-reasoning (FC) | xAI | 96.3 | 1/1 | Berkeley Function Calling Leaderboard |
#7 | Claude-Haiku-4-5-20251001 (FC) | Anthropic | 95.4 | 1/1 | Berkeley Function Calling Leaderboard |
#8 | mistral-medium-2508 | Mistral AI | 95.2 | 1/1 | Galileo Agent Leaderboard v2 |
#9 | Gemini-3-Pro-Preview (FC) | 94.4 | 1/1 | Berkeley Function Calling Leaderboard |
#10 | o3-2025-04-16 (Prompt) | OpenAI | 93.5 | 1/1 | Berkeley Function Calling Leaderboard |
#11 | Grok-4-0709 (Prompt) | xAI | 92.6 | 1/1 | Berkeley Function Calling Leaderboard |
#12 | Grok-4-0709 (FC) | xAI | 91.7 | 1/1 | Berkeley Function Calling Leaderboard |
#13 | Moonshotai-Kimi-K2-Instruct (FC) | moonshotai | 90.7 | 1/1 | Berkeley Function Calling Leaderboard |
#14 | gpt-4.1-mini-2025-04-14 | OpenAI | 90.5 | 1/1 | Galileo Agent Leaderboard v2 |
#15 | Grok-4-1-fast-non-reasoning (FC) | xAI | 89.8 | 1/1 | Berkeley Function Calling Leaderboard |
#16 | Command A Reasoning (FC) | cohere | 88.9 | 1/1 | Berkeley Function Calling Leaderboard |
#17 | DeepSeek-V3.2-Exp (Prompt + Thinking) | DeepSeek | 88.0 | 1/1 | Berkeley Function Calling Leaderboard |
#18 | Gemini-2.5-Flash (FC) | 87.0 | 1/1 | Berkeley Function Calling Leaderboard |
#19 | GPT-5.2-2025-12-11 (FC) | OpenAI | 86.1 | 1/1 | Berkeley Function Calling Leaderboard |
#20 | claude-sonnet-4-20250514 | Anthropic | 85.7 | 1/1 | Galileo Agent Leaderboard v2 |
#21 | GPT-5-mini-2025-08-07 (FC) | OpenAI | 85.2 | 1/1 | Berkeley Function Calling Leaderboard |
#22 | xLAM-2-32b-fc-r (FC) | salesforce | 84.3 | 1/1 | Berkeley Function Calling Leaderboard |
#23 | DeepSeek-V3.2-Exp (FC) | DeepSeek | 83.3 | 1/1 | Berkeley Function Calling Leaderboard |
#24 | GPT-4.1-2025-04-14 (FC) | OpenAI | 82.4 | 1/1 | Berkeley Function Calling Leaderboard |
#25 | o4-mini-2025-04-16 (FC) | OpenAI | 81.5 | 1/1 | Berkeley Function Calling Leaderboard |
#26 | kimi-k2-instruct | moonshot-ai | 81.0 | 1/1 | Galileo Agent Leaderboard v2 |
#27 | xLAM-2-70b-fc-r (FC) | salesforce | 80.6 | 1/1 | Berkeley Function Calling Leaderboard |
#28 | Qwen3-235B-A22B-Instruct-2507 (Prompt) | qwen | 79.6 | 1/1 | Berkeley Function Calling Leaderboard |
#29 | GPT-5-nano-2025-08-07 (FC) | OpenAI | 78.7 | 1/1 | Berkeley Function Calling Leaderboard |
#30 | Nanbeige4-3B-Thinking-2511 (FC) | nanbeige | 77.8 | 1/1 | Berkeley Function Calling Leaderboard |
#31 | Gemini-2.5-Flash (Prompt) | 76.9 | 1/1 | Berkeley Function Calling Leaderboard |
#32 | qwen3-235b-a22b-instruct-2507 | Alibaba | 76.2 | 1/1 | Galileo Agent Leaderboard v2 |
#33 | GPT-4.1-mini-2025-04-14 (FC) | OpenAI | 75.9 | 1/1 | Berkeley Function Calling Leaderboard |
#34 | o4-mini-2025-04-16 (Prompt) | OpenAI | 75.0 | 1/1 | Berkeley Function Calling Leaderboard |
#35 | Qwen3-32B (FC) | qwen | 74.1 | 1/1 | Berkeley Function Calling Leaderboard |
#36 | o3-2025-04-16 (FC) | OpenAI | 73.1 | 1/1 | Berkeley Function Calling Leaderboard |
#37 | Qwen3-235B-A22B-Instruct-2507 (FC) | qwen | 72.2 | 1/1 | Berkeley Function Calling Leaderboard |
#38 | qwen2.5-72b-instruct | Alibaba | 71.4 | 1/1 | Galileo Agent Leaderboard v2 |
#39 | Nanbeige3.5-Pro-Thinking (FC) | nanbeige | 71.3 | 1/1 | Berkeley Function Calling Leaderboard |
#40 | Qwen3-32B (Prompt) | qwen | 70.4 | 1/1 | Berkeley Function Calling Leaderboard |
#41 | xLAM-2-8b-fc-r (FC) | salesforce | 69.4 | 1/1 | Berkeley Function Calling Leaderboard |
#42 | Command A (FC) | cohere | 68.5 | 1/1 | Berkeley Function Calling Leaderboard |
#43 | BitAgent-Bounty-8B | bittensor | 67.6 | 1/1 | Berkeley Function Calling Leaderboard |
#44 | Arch-Agent-32B | katanemo | 66.7 | 1/1 | Berkeley Function Calling Leaderboard |
#45 | gemini-2.5-flash-lite | 66.7 | 1/1 | Galileo Agent Leaderboard v2 |
#46 | GPT-5.2-2025-12-11 (Prompt) | OpenAI | 65.7 | 1/1 | Berkeley Function Calling Leaderboard |
#47 | Qwen3-8B (FC) | qwen | 64.8 | 1/1 | Berkeley Function Calling Leaderboard |
#48 | ToolACE-2-8B (FC) | huawei-noah-ustc | 63.9 | 1/1 | Berkeley Function Calling Leaderboard |
#49 | Qwen3-30B-A3B-Instruct-2507 (FC) | qwen | 63.0 | 1/1 | Berkeley Function Calling Leaderboard |
#50 | xLAM-2-3b-fc-r (FC) | salesforce | 62.0 | 1/1 | Berkeley Function Calling Leaderboard |
#51 | glm-4.5-air | zai | 61.9 | 1/1 | Galileo Agent Leaderboard v2 |
#52 | Qwen3-14B (FC) | qwen | 61.1 | 1/1 | Berkeley Function Calling Leaderboard |
#53 | Qwen3-8B (Prompt) | qwen | 60.2 | 1/1 | Berkeley Function Calling Leaderboard |
#54 | GPT-4.1-2025-04-14 (Prompt) | OpenAI | 59.3 | 1/1 | Berkeley Function Calling Leaderboard |
#55 | mistral-large-2411 (FC) | mistral-ai | 58.3 | 1/1 | Berkeley Function Calling Leaderboard |
#56 | Qwen3-14B (Prompt) | qwen | 57.4 | 1/1 | Berkeley Function Calling Leaderboard |
#57 | gemini-2.5-pro | 57.1 | 1/1 | Galileo Agent Leaderboard v2 |
#58 | Mistral-Medium-2505 | mistral-ai | 56.5 | 1/1 | Berkeley Function Calling Leaderboard |
#59 | Mistral-Medium-2505 (FC) | mistral-ai | 55.6 | 1/1 | Berkeley Function Calling Leaderboard |
#60 | Llama-4-Maverick-17B-128E-Instruct-FP8 (FC) | Meta | 54.6 | 1/1 | Berkeley Function Calling Leaderboard |
#61 | Mistral-small-2506 (FC) | mistral-ai | 53.7 | 1/1 | Berkeley Function Calling Leaderboard |
#62 | Gemini-2.5-Flash-Lite (FC) | 52.8 | 1/1 | Berkeley Function Calling Leaderboard |
#63 | grok-4-0709 | xAI | 52.4 | 1/1 | Galileo Agent Leaderboard v2 |
#64 | Qwen3-30B-A3B-Instruct-2507 (Prompt) | qwen | 51.9 | 1/1 | Berkeley Function Calling Leaderboard |
#65 | Qwen3-4B-Instruct-2507 (FC) | qwen | 50.9 | 1/1 | Berkeley Function Calling Leaderboard |
#66 | Qwen3-4B-Instruct-2507 (Prompt) | qwen | 50.0 | 1/1 | Berkeley Function Calling Leaderboard |
#67 | Arch-Agent-3B | katanemo | 49.1 | 1/1 | Berkeley Function Calling Leaderboard |
#68 | Claude-Opus-4-5-20251101 (Prompt) | Anthropic | 48.1 | 1/1 | Berkeley Function Calling Leaderboard |
#69 | DeepSeek-V3 | DeepSeek | 47.6 | 1/1 | Galileo Agent Leaderboard v2 |
#70 | GPT-4.1-nano-2025-04-14 (FC) | OpenAI | 47.2 | 1/1 | Berkeley Function Calling Leaderboard |
#71 | Mistral-Small-2506 (Prompt) | mistral-ai | 46.3 | 1/1 | Berkeley Function Calling Leaderboard |
#72 | Arch-Agent-1.5B | katanemo | 45.4 | 1/1 | Berkeley Function Calling Leaderboard |
#73 | Command R7B (FC) | cohere | 44.4 | 1/1 | Berkeley Function Calling Leaderboard |
#74 | Llama-3.3-70B-Instruct (FC) | Meta | 43.5 | 1/1 | Berkeley Function Calling Leaderboard |
#75 | gemini-2.5-flash | 42.9 | 1/1 | Galileo Agent Leaderboard v2 |
#76 | mistral-large-2411 (Prompt) | mistral-ai | 42.6 | 1/1 | Berkeley Function Calling Leaderboard |
#77 | Hammer2.1-7b (FC) | madeagents | 41.7 | 1/1 | Berkeley Function Calling Leaderboard |
#78 | xLAM-2-1b-fc-r (FC) | salesforce | 40.7 | 1/1 | Berkeley Function Calling Leaderboard |
#79 | Gemma-3-12b-it (Prompt) | 39.8 | 1/1 | Berkeley Function Calling Leaderboard |
#80 | GPT-4.1-mini-2025-04-14 (Prompt) | OpenAI | 38.9 | 1/1 | Berkeley Function Calling Leaderboard |
#81 | gpt-4.1-nano-2025-04-14 | OpenAI | 38.1 | 1/1 | Galileo Agent Leaderboard v2 |
#82 | Hammer2.1-3b (FC) | madeagents | 38.0 | 1/1 | Berkeley Function Calling Leaderboard |
#83 | Gemma-3-27b-it (Prompt) | 37.0 | 1/1 | Berkeley Function Calling Leaderboard |
#84 | Phi-4 (Prompt) | microsoft | 36.1 | 1/1 | Berkeley Function Calling Leaderboard |
#85 | Qwen3-1.7B (FC) | qwen | 35.2 | 1/1 | Berkeley Function Calling Leaderboard |
#86 | Llama-4-Scout-17B-16E-Instruct (FC) | Meta | 34.3 | 1/1 | Berkeley Function Calling Leaderboard |
#87 | Gemini-2.5-Flash-Lite (Prompt) | 33.3 | 1/1 | Berkeley Function Calling Leaderboard |
#88 | qwen3-235b-a22b-thinking-2507 | Alibaba | 33.3 | 1/1 | Galileo Agent Leaderboard v2 |
#89 | CoALM-70B | uiuc-oumi | 32.4 | 1/1 | Berkeley Function Calling Leaderboard |
#90 | Hammer2.1-1.5b (FC) | madeagents | 31.5 | 1/1 | Berkeley Function Calling Leaderboard |
#91 | palmyra-x-004 (FC) | writer | 30.6 | 1/1 | Berkeley Function Calling Leaderboard |
#92 | GPT-5-mini-2025-08-07 (Prompt) | OpenAI | 29.6 | 1/1 | Berkeley Function Calling Leaderboard |
#93 | Open-Mistral-Nemo-2407 (FC) | mistral-ai | 28.7 | 1/1 | Berkeley Function Calling Leaderboard |
#94 | magistral-medium-2506 | Mistral AI | 28.6 | 1/1 | Galileo Agent Leaderboard v2 |
#95 | GPT-5-nano-2025-08-07 (Prompt) | OpenAI | 27.8 | 1/1 | Berkeley Function Calling Leaderboard |
#96 | Amazon-Nova-2-Lite-v1:0 (FC) | amazon | 26.9 | 1/1 | Berkeley Function Calling Leaderboard |
#97 | Granite-3.1-8B-Instruct (FC) | ibm | 25.9 | 1/1 | Berkeley Function Calling Leaderboard |
#98 | Falcon3-10B-Instruct (FC) | tii-uae | 25.0 | 1/1 | Berkeley Function Calling Leaderboard |
#99 | Granite-3.2-8B-Instruct (FC) | ibm | 24.1 | 1/1 | Berkeley Function Calling Leaderboard |
#100 | nova-pro-v1 | amazon | 23.8 | 1/1 | Galileo Agent Leaderboard v2 |
#101 | CoALM-8B | uiuc-oumi | 23.1 | 1/1 | Berkeley Function Calling Leaderboard |
#102 | Llama-3.1-8B-Instruct (Prompt) | Meta | 22.2 | 1/1 | Berkeley Function Calling Leaderboard |
#103 | MiniCPM3-4B-FC (FC) | openbmb | 21.3 | 1/1 | Berkeley Function Calling Leaderboard |
#104 | Claude-Haiku-4-5-20251001 (Prompt) | Anthropic | 20.4 | 1/1 | Berkeley Function Calling Leaderboard |
#105 | Amazon-Nova-Pro-v1:0 (FC) | amazon | 19.4 | 1/1 | Berkeley Function Calling Leaderboard |
#106 | mistral-small-2506 | Mistral AI | 19.0 | 1/1 | Galileo Agent Leaderboard v2 |
#107 | Claude-Sonnet-4-5-20250929 (Prompt) | Anthropic | 18.5 | 1/1 | Berkeley Function Calling Leaderboard |
#108 | GPT-4.1-nano-2025-04-14 (Prompt) | OpenAI | 17.6 | 1/1 | Berkeley Function Calling Leaderboard |
#109 | Falcon3-7B-Instruct (FC) | tii-uae | 16.7 | 1/1 | Berkeley Function Calling Leaderboard |
#110 | Qwen3-0.6B (FC) | qwen | 15.7 | 1/1 | Berkeley Function Calling Leaderboard |
#111 | Granite-20b-FunctionCalling (FC) | ibm | 14.8 | 1/1 | Berkeley Function Calling Leaderboard |
#112 | llama-3.3-70b-instruct | Meta | 14.3 | 1/1 | Galileo Agent Leaderboard v2 |
#113 | Qwen3-0.6B (Prompt) | qwen | 13.9 | 1/1 | Berkeley Function Calling Leaderboard |
#114 | Amazon-Nova-Micro-v1:0 (FC) | amazon | 13.0 | 1/1 | Berkeley Function Calling Leaderboard |
#115 | RZN-T (Prompt) | phronetic-ai | 12.0 | 1/1 | Berkeley Function Calling Leaderboard |
#116 | MiniCPM3-4B (Prompt) | openbmb | 11.1 | 1/1 | Berkeley Function Calling Leaderboard |
#117 | Llama-3.2-3B-Instruct (FC) | Meta | 10.2 | 1/1 | Berkeley Function Calling Leaderboard |
#118 | caller | arcee | 9.5 | 1/1 | Galileo Agent Leaderboard v2 |
#119 | Bielik-11B-v2.3-Instruct (Prompt) | speakleash-ack-cyfronet-agh | 9.3 | 1/1 | Berkeley Function Calling Leaderboard |
#120 | Hammer2.1-0.5b (FC) | madeagents | 8.3 | 1/1 | Berkeley Function Calling Leaderboard |
#121 | Gemma-3-4b-it (Prompt) | 7.4 | 1/1 | Berkeley Function Calling Leaderboard |
#122 | Open-Mistral-Nemo-2407 (Prompt) | mistral-ai | 6.5 | 1/1 | Berkeley Function Calling Leaderboard |
#123 | Granite-4.0-350m (FC) | ibm | 5.6 | 1/1 | Berkeley Function Calling Leaderboard |
#124 | magistral-small-2506 | Mistral AI | 4.8 | 1/1 | Galileo Agent Leaderboard v2 |
#125 | Falcon3-3B-Instruct (FC) | tii-uae | 4.6 | 1/1 | Berkeley Function Calling Leaderboard |
#126 | Ministral-8B-Instruct-2410 (FC) | mistral-ai | 3.7 | 1/1 | Berkeley Function Calling Leaderboard |
#127 | Falcon3-1B-Instruct (FC) | tii-uae | 2.8 | 1/1 | Berkeley Function Calling Leaderboard |
#128 | Llama-3.2-1B-Instruct (FC) | Meta | 1.9 | 1/1 | Berkeley Function Calling Leaderboard |
#129 | Llama-3.1-Nemotron-Ultra-253B-v1 (FC) | nvidia | 0.9 | 1/1 | Berkeley Function Calling Leaderboard |
#130 | Gemma-3-1b-it (Prompt) | 0.0 | 1/1 | Berkeley Function Calling Leaderboard |
#131 | nova-lite-v1 | amazon | 0.0 | 1/1 | Galileo Agent Leaderboard v2 |