Model comparisons - 119 pairs
AI model comparisons
119 hand-picked pairs from the model registry, each with first-party API prices, context windows, worked cost examples, setup guides and benchmark rows shown only where both models ran under the same setting. Pairs are chosen by three rules, listed below, rather than every possible combination.
Same provider, successor
31 pairs - input price per 1M tokens
Each model against the one it replaced in the same line, so you can see what changed in price, context and output limits.
- Claude Fable 5 vs Claude Fable 5.1$10 / $10
- Claude Haiku 4.5 vs Claude Haiku 5.5$1 / $0.10
- Claude Mythos 5 vs Claude Mythos 5.1$10 / $10
- Claude Opus 4.7 vs Claude Opus 4.8$5 / $5
- Claude Opus 4.8 vs Claude Opus 5$5 / $5
- Claude Opus 5 vs Claude Opus 5.5$5 / $4
- Claude Sonnet 4.6 vs Claude Sonnet 5$3 / $2
- Claude Sonnet 5 vs Claude Sonnet 5.5$2 / $2
- Gemini 2.5 Pro vs Gemini 3.1 Pro Preview$1.25 / $2
- Gemini 3.1 Flash-Lite vs Gemini 3.5 Flash-Lite$0.25 / $0.30
- Gemini 3.1 Pro Preview vs Gemini 4 Argon$2 / $2
- Gemini 3.6 Flash vs Gemini 3 Flash Preview$0.75 / $0.50
- Gemini 3.6 Flash vs Gemini 3.8 Flash$0.75 / $0.75
- GLM-5.1 vs GLM-5.2$1.40 / $1.40
- GLM-5.2 vs GLM-5.3$1.40 / $1.40
- GPT-5.4 vs GPT-5.5$2.50 / $5
- GPT-5.4 Mini vs GPT-5 Mini$0.75 / $0.25
- GPT-5.4 nano vs GPT-5 nano$0.20 / $0.05
- GPT-5.5 vs GPT-5.6 Sol$5 / $4
- GPT-5.6 Luna vs GPT-6 Luna$0.20 / $0.10
- GPT-5.6 Sol vs GPT-6 Sol$4 / $2
- GPT-6.1 Sol vs GPT-6 Sol$2 / $2
- Grok 4.5 vs Grok 4.6$2 / $2
- Grok 4.6 vs Grok 4.7$2 / $2
- Kimi K2.6 vs Kimi K3$0.95 / $3
- MiniMax-M2.7 vs MiniMax-M3$0.30 / $0.30
- Mistral Large 3 vs Mistral Large 4$0.50 / $1.36
- Qwen3.6-27B vs Qwen3.8-27B$0.60 / $0.50
- Qwen3.6-Plus vs Qwen3.7-Plus$0.50 / $0.40
- Qwen3.7-Flash vs Qwen3.8-Flash$0.03 / $0.15
- Qwen3.7-Max vs Qwen3.8-Max$2.50 / $2
Similar price, same tier
76 pairs - input price per 1M tokens
Models from different providers whose input prices are within 2x of each other and that sit in the same price tier.
- Claude Fable 5.1 vs GPT-6 Astra$10 / $10
- Claude Haiku 5.5 vs GLM-5.3-Flash$0.10 / $0.15
- Claude Haiku 5.5 vs GPT-6 Luna$0.10 / $0.10
- Claude Haiku 5.5 vs Mistral Small 4$0.10 / $0.15
- Claude Haiku 5.5 vs Qwen3.8-Flash$0.10 / $0.15
- Claude Opus 5.5 vs Gemini 4 Argon$4 / $2
- Claude Opus 5.5 vs GPT-6.1 Sol$4 / $2
- Claude Opus 5.5 vs Grok 4.7$4 / $2
- Claude Opus 5.5 vs Kimi K3$4 / $3
- Claude Opus 5.5 vs Qwen3.8-2.4T-A95B$4 / $2
- Claude Opus 5.5 vs Qwen3.8-Max$4 / $2
- Claude Sonnet 5.5 vs Gemini 4 Argon$2 / $2
- Claude Sonnet 5.5 vs GPT-6.1 Sol$2 / $2
- Claude Sonnet 5.5 vs Grok 4.7$2 / $2
- Claude Sonnet 5.5 vs Kimi K3$2 / $3
- Claude Sonnet 5.5 vs Qwen3.8-2.4T-A95B$2 / $2
- Claude Sonnet 5.5 vs Qwen3.8-Max$2 / $2
- DeepSeek V4 Pro vs Gemini 3.8 Flash$1.32 / $0.75
- DeepSeek V4 Pro vs GLM-5.3$1.32 / $1.40
- DeepSeek V4 Pro vs GPT-5.4 Mini$1.32 / $0.75
- DeepSeek V4 Pro vs Grok Build 0.1$1.32 / $1
- DeepSeek V4 Pro vs Kimi K2.7 Code$1.32 / $0.95
- DeepSeek V4 Pro vs Mistral Large 4$1.32 / $1.36
- DeepSeek V4 Pro vs Mistral Medium 3.5$1.32 / $1.50
- DeepSeek V4.1 Flash vs Gemini 3.5 Flash-Lite$0.30 / $0.30
- DeepSeek V4.1 Flash vs GLM-5.3-Flash$0.30 / $0.15
- DeepSeek V4.1 Flash vs MiniMax-M3$0.30 / $0.30
- DeepSeek V4.1 Flash vs Mistral Small 4$0.30 / $0.15
- DeepSeek V4.1 Flash vs Qwen3.8-Flash$0.30 / $0.15
- Gemini 3.5 Flash-Lite vs GLM-5.3-Flash$0.30 / $0.15
- Gemini 3.5 Flash-Lite vs MiniMax-M3$0.30 / $0.30
- Gemini 3.5 Flash-Lite vs Mistral Small 4$0.30 / $0.15
- Gemini 3.5 Flash-Lite vs Qwen3.8-Flash$0.30 / $0.15
- Gemini 3.8 Flash vs GLM-5.3$0.75 / $1.40
- Gemini 3.8 Flash vs GPT-5.4 Mini$0.75 / $0.75
- Gemini 3.8 Flash vs Grok Build 0.1$0.75 / $1
- Gemini 3.8 Flash vs Kimi K2.7 Code$0.75 / $0.95
- Gemini 3.8 Flash vs Mistral Large 4$0.75 / $1.36
- Gemini 3.8 Flash vs Mistral Medium 3.5$0.75 / $1.50
- Gemini 4 Argon vs GPT-6.1 Sol$2 / $2
- Gemini 4 Argon vs Grok 4.7$2 / $2
- Gemini 4 Argon vs Kimi K3$2 / $3
- Gemini 4 Argon vs Qwen3.8-2.4T-A95B$2 / $2
- Gemini 4 Argon vs Qwen3.8-Max$2 / $2
- GLM-5.3 vs GPT-5.4 Mini$1.40 / $0.75
- GLM-5.3 vs Grok Build 0.1$1.40 / $1
- GLM-5.3 vs Kimi K2.7 Code$1.40 / $0.95
- GLM-5.3 vs Mistral Large 4$1.40 / $1.36
- GLM-5.3 vs Mistral Medium 3.5$1.40 / $1.50
- GLM-5.3-Flash vs GPT-6 Luna$0.15 / $0.10
- GLM-5.3-Flash vs MiniMax-M3$0.15 / $0.30
- GLM-5.3-Flash vs Mistral Small 4$0.15 / $0.15
- GLM-5.3-Flash vs Qwen3.8-Flash$0.15 / $0.15
- GPT-5.4 Mini vs Grok Build 0.1$0.75 / $1
- GPT-5.4 Mini vs Kimi K2.7 Code$0.75 / $0.95
- GPT-5.4 Mini vs Mistral Large 4$0.75 / $1.36
- GPT-5.4 Mini vs Mistral Medium 3.5$0.75 / $1.50
- GPT-6 Luna vs Mistral Small 4$0.10 / $0.15
- GPT-6 Luna vs Qwen3.8-Flash$0.10 / $0.15
- GPT-6.1 Sol vs Grok 4.7$2 / $2
- GPT-6.1 Sol vs Kimi K3$2 / $3
- GPT-6.1 Sol vs Qwen3.8-2.4T-A95B$2 / $2
- GPT-6.1 Sol vs Qwen3.8-Max$2 / $2
- Grok 4.7 vs Kimi K3$2 / $3
- Grok 4.7 vs Qwen3.8-2.4T-A95B$2 / $2
- Grok 4.7 vs Qwen3.8-Max$2 / $2
- Grok Build 0.1 vs Kimi K2.7 Code$1 / $0.95
- Grok Build 0.1 vs Mistral Large 4$1 / $1.36
- Grok Build 0.1 vs Mistral Medium 3.5$1 / $1.50
- Kimi K2.7 Code vs Mistral Large 4$0.95 / $1.36
- Kimi K2.7 Code vs Mistral Medium 3.5$0.95 / $1.50
- Kimi K3 vs Qwen3.8-2.4T-A95B$3 / $2
- Kimi K3 vs Qwen3.8-Max$3 / $2
- MiniMax-M3 vs Mistral Small 4$0.30 / $0.15
- MiniMax-M3 vs Qwen3.8-Flash$0.30 / $0.15
- Mistral Small 4 vs Qwen3.8-Flash$0.15 / $0.15
Open weights vs closed
12 pairs - input price per 1M tokens
Leading open-weights models against closed models of a comparable class, frontier with frontier and small with small.
- Claude Haiku 5.5 vs DeepSeek V4.1 Flash$0.10 / $0.30
- Claude Haiku 5.5 vs MiniMax-M3$0.10 / $0.30
- Claude Opus 5.5 vs DeepSeek V4 Pro$4 / $1.32
- Claude Opus 5.5 vs GLM-5.3$4 / $1.40
- Claude Sonnet 5.5 vs DeepSeek V4 Pro$2 / $1.32
- Claude Sonnet 5.5 vs GLM-5.3$2 / $1.40
- DeepSeek V4 Pro vs Gemini 4 Argon$1.32 / $2
- DeepSeek V4 Pro vs GPT-6.1 Sol$1.32 / $2
- DeepSeek V4.1 Flash vs GPT-6 Luna$0.30 / $0.10
- Gemini 4 Argon vs GLM-5.3$2 / $1.40
- GLM-5.3 vs GPT-6.1 Sol$1.40 / $2
- GPT-6 Luna vs MiniMax-M3$0.10 / $0.30
Newsletter
Get model price changes as they ship
One email per week. New model releases, price cuts and what they mean for your bill.