DeepSeek V4 Pro
DeepSeek open 2026-04-24
Killed
DeepSeek V4 Pro is killed. The executioner, Claude Haiku 5.5, costs $0.20 and scores 43.4: no more expensive, no weaker. Choosing DeepSeek V4 Pro means paying more for nothing.
- Score
- 36.0
- Artificial Analysis Intelligence Index
- Best rank
- #26
- Arena — Agent
- Price
- $0.54
- Blended Price, USD per 1M tokens (3:1 input:output)
- Best effort
- max
- Artificial Analysis LLM Leaderboard
Not among the SOTA models right now, so this page keeps to the verdict and the scores.
Scores on each board Each rank is the board’s own: the rank it publishes, or the model’s place in the board’s own order where it publishes none. The bar shows how much of that board the model ranks ahead of. Scores from different boards cannot be compared.
- Artificial Analysis LLM LeaderboardArtificial Analysis Intelligence Index36.0No. 38 of 375
- Arena — AgentOverall0.013No. 26 of 52
- Artificial Analysis Coding AgentsArtificial Analysis Coding Agent Index0.431No. 20 of 24
- Epoch AI Benchmarking HubFrontierMath Tiers 1–364.6%No. 32 of 82
- Arena — CodeOverall1582No. 32 of 142
- Arena — TextOverall (style control)1465No. 55 of 414
Compare with the executioner, Claude Haiku 5.5 → See the current SOTA models →
Details
- Family
- DeepSeek → DeepSeek V4
- Blended Price, USD per 1M tokens (3:1 input:output)
- $0.54
- Sources
- 6 boards ·15 scores ·11 settings
- Also written as
-
DeepSeek V4 Prodeepseek-ai/DeepSeek-V4-Prodeepseek-v4-pro