DeepSeek V3 (Dec '24) vs MiMo-V2.5-Pro

Which is better for the money: DeepSeek V3 (Dec '24) by DeepSeek or MiMo-V2.5-Pro by Xiaomi? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-07-17.

Trade-off
DeepSeekDeepSeek V3 (Dec '24)On the frontier
Intelligence
14.20
Total cost
$19.8
XiaomiMiMo-V2.5-ProOn the frontier
Intelligence
42.20
Total cost
$98.5
Verdict

Depends on your budget

MiMo-V2.5-Pro scores 28.0 points higher (42.2 vs 14.2) but costs 5.0× as much to run the benchmark ($98.5 vs $19.8; mid-range vs budget cost classes). They are not alternatives at one budget: pick MiMo-V2.5-Pro when you need intelligence above 14.2, and DeepSeek V3 (Dec '24) when 14.2 is enough.

MetricDeepSeek V3 (Dec '24)MiMo-V2.5-ProDifference
Intelligence (higher is better)14.242.2+197.2% MiMo-V2.5-Pro is 3.0× more intelligent
Total AA cost (lower is better)$19.84$98.47+396.3% MiMo-V2.5-Pro is 5.0× more expensive
End-to-end time (lower is better)Not reported46.21s— not comparable
Price per 1M tokens (in / out)$0.36 / $0.89$0.43 / $0.87
Input modalitiesNot reportedNot reported

MiMo-V2.5-Pro vs DeepSeek V3 (Dec '24): 197.2% more intelligence, 396.3% more cost to run the AA benchmark, and end-to-end time not fully reported.

Open this matchup in the interactive pickerDeepSeek V3 (Dec '24): what beats itMiMo-V2.5-Pro: what beats it
Today's Pareto frontier

Against the frontier

0204060$10$20$50$100$200$500$1k$2k$5k$10kCost to run the Artificial Analysis Intelligence Index (USD, log scale)Intelligence IndexClaude 4.5 Haiku (Reasoning) (Anthropic): intelligence 29.60, $564Claude 4.5 Sonnet (Reasoning) (Anthropic): intelligence 36.40, $1133Claude Opus 4.7 (Adaptive Reasoning, Max Effort) (Anthropic): intelligence 53.50, $3738Claude Opus 4.8 (Adaptive Reasoning, Max Effort) (Anthropic): intelligence 55.70, $3753Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) (Anthropic): intelligence 47.20, $3356Claude Sonnet 5 (Adaptive Reasoning, Max Effort) (Anthropic): intelligence 53.40, $4010Claude Sonnet 5 (Non-reasoning, High Effort) (Anthropic): intelligence 41.70, $427DeepSeek R1 (Jan '25) (DeepSeek): intelligence 18.50, $416DeepSeek V3 0324 (DeepSeek): intelligence 15.40, $132Gemini 2.5 Pro (Google): intelligence 25.80, $610Gemini 3.1 Flash-Lite (Google): intelligence 25.00, $93.7Gemini 3.1 Pro Preview (Google): intelligence 46.50, $815Gemini 3.5 Flash (high) (Google): intelligence 50.20, $1041Gemma 4 26B A4B (Reasoning) (Google): intelligence 25.70, $51.9Gemma 4 31B (Non-reasoning) (Google): intelligence 21.80, $29.0GLM-4.6 (Reasoning) (Z AI): intelligence 28.70, $313GLM-4.7 (Reasoning) (Z AI): intelligence 33.70, $603GLM-5.1 (Reasoning) (Z AI): intelligence 40.20, $659GLM-5.2 (max) (Z AI): intelligence 51.10, $925GPT-4.1 mini (OpenAI): intelligence 14.80, $74.7GPT-4.1 nano (OpenAI): intelligence 9.60, $48.7GPT-5 (high) (OpenAI): intelligence 34.70, $813GPT-5.4 (xhigh) (OpenAI): intelligence 51.40, $2132GPT-5.4 mini (xhigh) (OpenAI): intelligence 40.00, $1095GPT-5.5 (high) (OpenAI): intelligence 53.10, $1719GPT-5.5 (low) (OpenAI): intelligence 43.50, $375GPT-5.5 (medium) (OpenAI): intelligence 50.40, $936GPT-5.5 (Non-reasoning) (OpenAI): intelligence 35.40, $211GPT-5.5 (xhigh) (OpenAI): intelligence 54.80, $2778GPT-5.5 Instant (June 2026) (OpenAI): intelligence 28.90, $528GPT-5.6 Luna (low) (OpenAI): intelligence 33.30, $68.8GPT-5.6 Luna (max) (OpenAI): intelligence 51.20, $870GPT-5.6 Luna (medium) (OpenAI): intelligence 38.10, $106GPT-5.6 Luna (Non-reasoning) (OpenAI): intelligence 26.60, $51.4GPT-5.6 Luna (xhigh) (OpenAI): intelligence 49.10, $479GPT-5.6 Sol (Non-reasoning) (OpenAI): intelligence 41.20, $242GPT-5.6 Terra (high) (OpenAI): intelligence 49.00, $496GPT-5.6 Terra (low) (OpenAI): intelligence 40.50, $161GPT-5.6 Terra (max) (OpenAI): intelligence 55.00, $1754GPT-5.6 Terra (xhigh) (OpenAI): intelligence 51.60, $740gpt-oss-120b (high) (OpenAI): intelligence 23.80, $96.3gpt-oss-20b (high) (OpenAI): intelligence 14.90, $29.9Grok 4.3 (high) (SpaceXAI): intelligence 37.60, $288Grok 4.3 (Non-reasoning) (SpaceXAI): intelligence 24.80, $201Grok Build 0.1 0616 (SpaceXAI): intelligence 39.80, $375HyperNova 60B 2605 (Multiverse Computing): intelligence 17.80, $35.0Kimi K3 (Kimi): intelligence 57.10, $2691Llama 3.3 Instruct 70B (Meta): intelligence 9.40, $78.0Llama 4 Maverick (Meta): intelligence 14.30, $38.2Magistral Medium 1.2 (Mistral): intelligence 17.90, $886Ministral 3 14B (Mistral): intelligence 11.10, $120Ministral 3 3B (Mistral): intelligence 6.80, $91.8Ministral 3 8B (Mistral): intelligence 9.00, $113Mistral Large 3 (Mistral): intelligence 15.90, $71.0Mistral Medium 3.1 (Mistral): intelligence 14.70, $132Mistral Medium 3.5 (Mistral): intelligence 29.90, $1483Mistral Small 3.1 (Mistral): intelligence 14.70, $29.9Mistral Small 3.2 (Mistral): intelligence 10.60, $128Nemotron 3 Ultra 550B A55B (Reasoning) (NVIDIA): intelligence 37.80, $435Nova 2.0 Lite (high) (Amazon): intelligence 18.20, $635Nova 2.0 Pro Preview (low) (Amazon): intelligence 19.60, $244Nova 2.0 Pro Preview (medium) (Amazon): intelligence 21.80, $407Nova 2.0 Pro Preview (Non-reasoning) (Amazon): intelligence 14.40, $206NVIDIA Nemotron 3 Nano 30B A3B (Reasoning) (NVIDIA): intelligence 14.20, $38.9NVIDIA Nemotron 3 Super 120B A12B (Reasoning) (NVIDIA): intelligence 25.40, $254Qwen3 30B A3B 2507 (Reasoning) (Alibaba): intelligence 14.40, $152Qwen3 Coder Next (Alibaba): intelligence 21.10, $247Qwen3 Next 80B A3B (Reasoning) (Alibaba): intelligence 16.70, $401Qwen3.5 122B A10B (Non-reasoning) (Alibaba): intelligence 27.60, $216Qwen3.5 122B A10B (Reasoning) (Alibaba): intelligence 32.30, $447Qwen3.5 397B A17B (Reasoning) (Alibaba): intelligence 33.70, $528Qwen3.5 9B (Reasoning) (Alibaba): intelligence 21.40, $179Qwen3.6 27B (Non-reasoning) (Alibaba): intelligence 30.50, $356Qwen3.6 27B (Reasoning) (Alibaba): intelligence 37.10, $668Qwen3.6 35B A3B (Non-reasoning) (Alibaba): intelligence 24.20, $496Qwen3.6 35B A3B (Reasoning) (Alibaba): intelligence 31.60, $334Qwen3.6 Plus (Alibaba): intelligence 39.60, $526Qwen3.7 Max (Alibaba): intelligence 46.00, $1559Qwen3.7 Plus (Alibaba): intelligence 39.00, $303Ring-2.6-1T (InclusionAI): intelligence 30.60, $460Llama 4 Scout (Meta): intelligence 10.00, $11.5gpt-oss-120b (low) (OpenAI): intelligence 14.90, $24.2MiMo-V2.5 (Xiaomi): intelligence 37.20, $25.2DeepSeek V4 Flash (Reasoning, Max Effort) (DeepSeek): intelligence 40.30, $74.3DeepSeek V4 Pro (Reasoning, Max Effort) (DeepSeek): intelligence 44.30, $176MiniMax-M3 (MiniMax): intelligence 44.40, $204GPT-5.6 Terra (medium) (OpenAI): intelligence 45.60, $240GPT-5.6 Luna (high) (OpenAI): intelligence 46.10, $275GPT-5.6 Sol (low) (OpenAI): intelligence 49.40, $353Muse Spark 1.1 (xhigh) (Meta): intelligence 50.60, $548GPT-5.6 Sol (medium) (OpenAI): intelligence 53.60, $593Grok 4.5 (high) (SpaceXAI): intelligence 53.80, $602GPT-5.6 Sol (high) (OpenAI): intelligence 55.90, $956GPT-5.6 Sol (xhigh) (OpenAI): intelligence 57.70, $1543GPT-5.6 Sol (max) (OpenAI): intelligence 58.90, $2824Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) (Anthropic): intelligence 59.90, $5631MiMo-V2.5-Pro (Xiaomi): intelligence 42.20, $98.5DeepSeek V3 (Dec '24) (DeepSeek): intelligence 14.20, $19.8DeepSeek V3 (Dec '24)MiMo-V2.5-ProLlama 4 ScoutClaude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)
DeepSeek V3 (Dec '24)MiMo-V2.5-ProToday's Pareto frontierEvery other model
Matchup

Compare something else

Agents: Accept: text/markdown or ?format=md returns this page as Markdown. JSON: /api/v1/compare/deepseek-v3/mimo-v2-5-pro · Markdown: /api/v1/compare/deepseek-v3/mimo-v2-5-pro?format=md · llms.txt · agent skill