The arena · AI model review
DeepSeek-V4
by DeepSeek
Open-weights 1.6T MoE with 1M-token context and near-frontier agentic coding at $0.87 per 1M output tokens
$0.87/1M out
V4-Pro tier: $0.435/1M in ($0.003625 cache hit), $0.87/1M out; cheaper V4-Flash tier at $0.14/$0.28 ($0.0028 cache hit); the 75% launch discount became permanent pricing on 2026-05-22. The official mid-July 2026 release introduces peak/off-peak pricing, with listed rates doubling during Beijing peak hours.
DeepSeek
1M tokens (384K max output)
$0.435/1M in (cache hit $0.003625)
$0.87/1M out
text only
Yes
What is DeepSeek-V4?
Flagship open-weights MoE family from DeepSeek (V4-Pro 1.6T params/49B active, V4-Flash 284B/13B), released as a preview on April 24, 2026. Brings a 1M-token context window via sparse attention, thinking and non-thinking modes, and MIT-licensed weights on Hugging Face. Positioned as open-source SOTA on agentic coding benchmarks at a fraction of frontier API prices.
DeepSeek-V4 pros & cons
Pros
- 1M-token context window (8x the 128K of V3.2) with up to 384K output tokens, standard on the official API
- Aggressive pricing: $0.435/$0.87 per 1M tokens (V4-Pro), roughly 28.7x cheaper per output token than Claude Opus 4.8; cache-hit input drops to $0.003625/1M (over 99% discount)
- MIT-licensed open weights for both V4-Pro and V4-Flash on Hugging Face: commercial use, fine-tuning and redistribution allowed
- Open-source SOTA on agentic coding: 80.6 on SWE-bench Verified (Think Max config), tied with Gemini 3.1 Pro, plus Codeforces rating 3206 (~rank 23 vs humans)
- Ranks #3 of 93 on the Artificial Analysis Intelligence Index (score 44), well above the 25 average
- Sparse-attention stack cuts 1M-context inference to 27% of V3.2's FLOPs and 10% of its KV cache
Cons
- Intermittent malformed tool calls: function calls sometimes emitted as plain text in content instead of the tool_calls field (GitHub issue deepseek-ai #1244)
- Thinking mode breaks long multi-turn tool-call chains with 400 errors in agent frameworks (OpenClaw issue #72044, fix still incomplete)
- Developers report it fabricating nonexistent APIs in custom codebases and acting on hallucinated user input in agent loops
- Very verbose (180M eval output tokens vs 95M median) and mid-pack speed at 54.6 tok/s (#39/93), which erodes the low per-token price in practice
- Text only (no vision or audio) and still a preview: the official release planned for mid-July 2026 adds peak-hour pricing that doubles listed API rates during Beijing business hours
The arena’s verdict
Choose DeepSeek-V4 if you want near-frontier reasoning and agentic coding at 3x to nearly 30x below Claude Opus or GPT-5.5 pricing, or if MIT-licensed weights for self-hosting and fine-tuning matter to you. It is a decisive upgrade over V3.2: 8x longer context, far cheaper long-context inference and stronger coding, and the legacy deepseek-chat/reasoner endpoints are deprecated on July 24, 2026 anyway. Avoid it for production agents that depend on rock-solid multi-turn tool calling, where users still report malformed tool calls and fabricated APIs, and for any vision or audio work since it is text only. Latency-sensitive apps should also test first, as its verbosity and mid-pack 54.6 tok/s output speed offset some of the cost advantage, and budget for the peak-hour price doubling arriving with the official mid-July release.
Thumbs up or thumbs down
Cast your verdict
Would you recommend DeepSeek-V4, or warn the crowd away?
Top DeepSeek-V4 alternatives
All alternativesAnthropic's fastest model: about 90% of Sonnet 4.5's coding skill at $1/$5 per 1M tokens, 200K context.
Anthropic's April 2026 Opus: 87.6% SWE-bench Verified, 1M context, high-res vision, now behind Opus 4.8
Anthropic's flagship Opus-tier model for long-horizon agentic coding; 1M context at $5/$25 per 1M tokens.
Compare DeepSeek-V4 head-to-head
What the crowd says
“Tool calling is flaky. Function calls sometimes land as plain text instead of the tool_calls field, and thinking mode 400s on long multi-turn chains. Not agent-ready yet.”
“MIT license on both Pro and Flash weights is the real story. Fine-tune, redistribute, ship commercially, no lawyer needed. Plus 384K output tokens for long-doc generation.”
“MIT weights, 1M context, and output tokens roughly 29x cheaper than Opus 4.8. Cache hits make input basically free. Moved my bulk pipelines over and the bill collapsed.”
DeepSeek-V4: frequently asked questions
How much does DeepSeek-V4 cost per 1M tokens?
DeepSeek-V4 costs $0.435/1M in (cache hit $0.003625) per 1M input tokens and $0.87/1M out per 1M output tokens. V4-Pro tier: $0.435/1M in ($0.003625 cache hit), $0.87/1M out; cheaper V4-Flash tier at $0.14/$0.28 ($0.0028 cache hit); the 75% launch discount became permanent pricing on 2026-05-22. The official mid-July 2026 release introduces peak/off-peak pricing, with listed rates doubling during Beijing peak hours.