Peak-valley pricing · 1M context · 80.6% SWE-bench · Migrate before July 24
After three months of preview access, DeepSeek V4 officially hit general availability on July 20, 2026. If you are asking is DeepSeek V4 better than GPT-5.6 or still calling deepseek-chat in production, this developer-focused guide covers the full timeline, V4-Pro/Flash architecture, benchmark tables, peak-valley pricing, an honest comparison with Claude Fable 5, and a six-step migration runbook before the July 24 deadline. GA brings agent/math/code upgrades plus time-based pricing — SWE-bench Verified hits 80.6% (open-weight record) at $0.87/M output off-peak. Data as of July 20, 2026.
The preview dropped April 24, 2026. Today's GA is not a new architecture — it graduates the preview into production with stability, optimization, and commercial pricing discipline.
| Date | Milestone |
|---|---|
| Apr 24, 2026 | Preview + MIT open source: V4-Pro (1.6T) and V4-Flash (284B) |
| May 2026 | Production-tuned versions; API generally available |
| Jun 2026 | V4-Pro output permanently cut 75% to $0.87/M tokens |
| Jun 29, 2026 | Email to all API users: GA mid-July + first peak-valley pricing disclosure |
| Jul 19, 2026 | Gray-release testers; media reports imminent full launch |
| Jul 20, 2026 | GA live globally |
| Jul 24, 2026 | deepseek-chat and deepseek-reasoner permanently disabled (15:59 UTC) |
Five pain points developers face right now:
Legacy endpoints dying in 4 days: deepseek-chat and deepseek-reasoner stop working July 24.
Peak pricing complexity: Weekday business-hour rates double — 24/7 pipelines need scheduling.
Closed-model costs: Claude Fable 5 at ~$50/M output and GPT-5.6 Sol at ~$15/M drain budgets at scale.
1M context on paper only: Most models choke on KV cache memory at long context.
Data sovereignty: Closed APIs cannot be self-hosted for compliance-sensitive workloads.
| Spec | V4-Pro | V4-Flash |
|---|---|---|
| Total parameters | 1.6 trillion | 284 billion |
| Active per token | 49B (3%) | 13B (4.6%) |
| Layers | 61 | 43 |
| Context window | 1,000,000 tokens | 1,000,000 tokens |
| Max output | 384K | 384K |
| Precision | FP4 (MoE experts) + FP8 | Same |
| Training data | 33T+ tokens | 32T+ tokens |
| License | MIT | MIT |
V4 replaced MLA from V2/V3 with a two-track system. Compressed Sparse Attention (CSA) compresses KV 4x via softmax-gated pooling, uses an FP4 lightning indexer (top-1024 Pro / top-512 Flash), plus a 128-token sliding window. Heavily Compressed Attention (HCA) compresses 128x then runs dense global attention. Layers alternate CSA and HCA.
At 1M tokens: V4-Pro uses 27% of V3.2 inference FLOPs; KV cache drops to 10% of V3.2 memory (Flash: 7%).
Manifold-Constrained Hyper-Connections (mHC) replace standard residuals with a 4-channel stream governed by a doubly stochastic matrix — stable gradients through 61 layers. Muon (not AdamW) uses Newton-Schulz orthogonalization for faster, stabler training.
| Mode | Use case |
|---|---|
| Non-think | Fast routing, classification, simple Q&A |
| Think High | Debugging, medium complexity |
| Think Max | Complex math, multi-step agents (384K+ context) |
Recommended sampling: temperature=1.0, top_p=1.0 for all modes.
| Benchmark | V4-Pro | Claude Fable 5 | GPT-5.6 Ultra | Opus 4.8 |
|---|---|---|---|---|
| SWE-bench Verified | 80.6% open record | 96.0% | N/A | ~69% |
| SWE-bench Pro | 55.4% | 80.3% | 78.1% | 69.2% |
| LiveCodeBench | 93.5% | 88.1% | 87.4% | 83.2% |
| Codeforces Elo | 3,206 | — | — | — |
| Terminal-Bench 2.1 | 83.9% | 88.0% | 85.1% | 82.7% |
Artificial Analysis Strategy & Ops index: Fable 5 scores 50 at $3.48/task; V4-Pro scores 38 at $0.03/task — 116x cheaper for a 24% performance gap. V4-Flash stays under $0.04/task across all six index categories.
| Dimension | V4-Pro | GPT-5.6 Sol | Claude Fable 5 |
|---|---|---|---|
| Open / self-hostable | Yes (MIT) | No | No |
| 1M context | Yes | Not confirmed | Yes |
| Best coding (hardest repos) | Second tier | Second tier | Leads SWE-bench Pro |
| Algorithms / math | LiveCodeBench leader | Strong | — |
| Output (off-peak) | $0.87/M | ~$15/M | ~$50/M |
| Output (peak) | $1.74/M | — | — |
| Data sovereignty | Self-host possible | No | No |
Peak hours (Beijing Time): Weekdays 09:00–12:00 and 14:00–18:00 — all rates double.
| Model | Item | Off-Peak | Peak |
|---|---|---|---|
| V4-Pro | Input (cache hit) | $0.0035/M | 2x |
| Input (cache miss) | $0.435/M | $0.87/M | |
| Output | $0.87/M | $1.74/M | |
| V4-Flash | Input (cache hit) | $0.0028/M | 2x |
| Input (cache miss) | $0.14/M | $0.28/M | |
| Output | $0.28/M | $0.56/M |
Even at peak, V4-Pro output is 8.6x cheaper than Claude Opus 4.8 ($15/M) and GPT-5.6 Sol (~$15/M).
Cost tips: Schedule batch jobs off-peak; maximize prompt cache hits; route simple queries to V4-Flash; monitor DeepSeek billing-change emails (24h notice).
Deadline: July 24, 2026, 15:59 UTC. Legacy names deepseek-chat and deepseek-reasoner will be permanently disabled.
| Old name | New equivalent |
|---|---|
deepseek-chat | deepseek-v4-flash (non-thinking) |
deepseek-reasoner | deepseek-v4-flash (thinking) or deepseek-v4-pro |
Search codebase: Grep for deepseek-chat and deepseek-reasoner including env vars and CI configs.
Pick target model: Flash for chat/routing; Pro for heavy reasoning.
Update OpenAI SDK: Change model only; base_url stays https://api.deepseek.com.
Enable thinking mode: Use extra_body with thinking config to replace reasoner behavior.
Test in staging: Validate core flows; Think Max needs 384K+ context window.
Ship before deadline: Anthropic SDK users swap model name only — same base URL.
response = client.chat.completions.create(
model="deepseek-v4-pro",
messages=[{"role": "user", "content": "Solve step by step..."}],
extra_body={"thinking": {"type": "enabled", "budget_tokens": 8000}}
)
The GA release does not need to beat Claude Fable 5 on every benchmark. It delivers the strongest open-weight coding model at 1/10 to 1/100 the cost of closed frontier APIs.
Self-hosting V4 or running long-context agents still needs reliable hardware — laptops cannot stay up 24/7 and VMs add performance and compliance risk. For production iOS CI/CD and AI agent automation, VpsMesh Mac Mini cloud rental is usually the better fit: bare-metal Apple Silicon pairs well with local inference stacks like ds4 + V4 Flash on Mac.
Sources: DeepSeek official docs, arXiv:2606.19348, HuggingFace, LLMReference, Artificial Analysis, MangoMind Blog, TechNode.
Weekday Beijing hours 09:00–12:00 and 14:00–18:00 double all rates. V4-Pro off-peak output is $0.87/M; peak is $1.74/M. See our Mac Mini rental pricing for complementary dev environment costs.
Before July 24, replace deepseek-chat with deepseek-v4-flash and deepseek-reasoner with Flash thinking mode or deepseek-v4-pro. Base URL unchanged.
Pro is the 1.6T flagship (49B active); Flash is 284B (13B active). Both support 1M context. Flash is cheaper ($0.28/M output) for routing; Pro for complex agents.
Fable 5 and GPT-5.6 lead on hardest benchmarks. V4-Pro is MIT-licensed, self-hostable, and among the cheapest frontier AI APIs in 2026 at $0.87/M off-peak output.
V4-Flash can run via engines like ds4 on high-end Macs (96GB+ unified memory). Full Pro needs cluster hardware. Setup details in our help center.
Agent, math, and code performance upgrades; peak-valley pricing; global GA availability; legacy model names retire July 24. Underlying MoE architecture unchanged.