July 31, 2026 · V4-Flash-0731 official · Post-training only · Harness teased · Pro still undated
Summary: On July 31, DeepSeek promoted deepseek-v4-flash to an official public API build (tag 0731) with the same 284B-parameter architecture as April's preview—gains came entirely from a fresh post-training pass. Agent scores now rival Claude Opus 4.8 at roughly 1/36 to 1/179 of the price (vendor list rates). Flagship V4-Pro and in-house agent framework Harness remain unreleased. This piece covers the timeline, pricing tables, architecture notes, Chinese open-weight rivalry, benchmark caveats, the "kill line" pricing narrative, and a five-step developer checklist.
Legacy API names are dead: deepseek-chat and deepseek-reasoner retired July 24—unmigrated production breaks immediately
"Official" ≠ new architecture: 0731 is a re-trained Flash, not a bigger model—wrong expectations skew routing decisions
Agent scores are Harness-dependent: Terminal Bench 2.0 figures used DeepSeek's unreleased minimal-mode Harness—don't port them blindly to Claude Code or Cursor
Pro and Harness still missing: Only Flash is official, API-only; app and web chat untouched; August 10–20 GA rumors are unconfirmed
| Date | Event |
|---|---|
| Apr 24, 2026 | V4 preview + MIT open weights: V4-Pro (1.6T/49B active) and V4-Flash (284B/13B active), 1M context |
| Jul 24, 2026 | Legacy aliases retired; all traffic routes to V4 family names |
| Jul 27, 2026 | Moonshot AI ships Kimi K3 open weights (2.8T total)—direct competitive pressure |
| Jul 31, 2026 | V4-Flash-0731 official API beta; weights on Hugging Face; changelog names Harness "coming soon"; API-only |
| As of Aug 5, 2026 | Official V4-Pro still "as soon as possible"; media August 10–20 window not confirmed by DeepSeek |
| Model | Status | Total / active | Context | Input $/M (miss / hit) | Output $/M | License |
|---|---|---|---|---|---|---|
| V4-Flash-0731 | Official | 284B / 13B | 1M | $0.14 / $0.0028 | $0.28 | MIT |
| V4-Pro | Preview only | 1.6T / 49B | 1M | $0.435 / $0.003625 | $0.87 | MIT |
| Kimi K3 | Open weights Jul 27 | 2.8T / ~104B est. | ~1.05M | $3.00 / $0.30 | $15.00 | Modified MIT |
| GLM-5.2 | Open Jun 2026 | ~744B / ~40B | 1M | Not verified here | MIT | |
| Qwen3.8-Max | API GA Aug 2 | 2.4T / 95B | 1M | $2.00 / ~$0.17–0.25 | $6.00 | Weights pending |
DeepSeek announced a future 2× peak-hour surcharge (Beijing 9–12, 14–18) with no effective date yet. Figures as of August 5, 2026.
V4-Flash-0731 is identical in size and structure to April's preview. DeepSeek attributes agent benchmark jumps entirely to re-run post-training—a 284B/13B model now beats the family's 1.6T/49B preview on multiple agentic tasks, underscoring that late-2026 competition is as much about training data and methods as raw parameter count.
The technical report describes CSA + HCA hybrid attention ("DSA"), manifold-constrained hyper-connections (mHC), and the Muon optimizer. Vendor claims at 1M tokens: V4-Pro needs 27% of V3.2 per-token FLOPs and 10% of KV cache—pending independent reproduction.
July 31 marked the first official mention of DeepSeek Harness—file I/O, tool calls, engineering workflows—DeepSeek's answer to Claude Code. Every published agent score (Terminal Bench 2.0, Toolathlon, etc.) used Harness minimal mode (not yet public). The changelog warns scores are "extremely sensitive to harness choice."
| Model | Lab | Total params | AA Intelligence Index | Cost / task (AA) |
|---|---|---|---|---|
| V4-Flash-0731 | DeepSeek | 284B | 50 | $0.03 |
| Kimi K3 | Moonshot | 2.8T | 57 | $0.86 |
| GLM-5.2 | Zhipu | ~744B | ~1 pt above Flash | Not verified |
| GPT-5.6 Sol | OpenAI | Undisclosed | 9+ pts above | $1.86 |
| Claude Fable 5 | Anthropic | Undisclosed | 9+ pts above | $3.15 |
On Artificial Analysis's independent index, Flash isn't the smartest—but per-task cost is ~1/29 of Kimi K3 and ~1/105 of Claude Fable 5. DeepSeek optimizes for "good-enough intelligence at unmatched price," which also explains seven weeks atop OpenRouter usage for the preview.
| Scenario | Lean toward | Why |
|---|---|---|
| High-volume agents / API batches | V4-Flash-0731 | Lowest cost; 0731 agent scores beat Pro preview |
| Max independent index | Kimi K3 or closed flagship | Higher AA score, much higher unit cost |
| Self-host + 1M context | V4-Flash MIT weights | Open license on Hugging Face |
| Waiting on Harness | Hold for Pro official | Framework unreleased; third-party agents still required |
Harness lock-in: Terminal Bench 2.0 at 82.7 (vs Pro preview 67.9) used unreleased Harness minimal mode—treat as vendor + framework, not portable truth
Usability reports: Overseas devs cite low input cache-hit rates and occasional safety-classifier timeouts
Pro/Harness dates: August 10–20 GA rumors are unnamed-source media—not DeepSeek confirmation
Funding rumors: ~$7.4B round / ~$48.7B valuation reports lack regulatory filings—background only
Chinese forums nicknamed founder Liang Wenfeng "Liang Empty Promise" when Pro slipped past mid-July; after 0731 over-delivered, "Liang the Sage" returned—a fast sentiment barometer.
Developers also use 斩杀线 (zhǎn shā xiàn)—"kill line": good-enough capability plus rock-bottom price sets a bar competitors must beat on both axes. OpenAI's reported 80% Luna price cut fits that framing. On July 31, Nvidia, Broadcom, and AMD barely moved—markets now treat "DeepSeek efficiency" as routine engineering, unlike the 2025 R1 chip selloff.
Point production at deepseek-v4-flash; retire legacy aliases
Separate vendor agent scores from AA independent index when routing
Schedule bulk jobs outside announced 2× peak windows (date TBD)
Validate agents in OpenClaw / Claude Code—don't rely on TB2.0 alone
Watch changelog for Harness and Pro—API and weights may diverge again
Yes—V4-Pro and V4-Flash, including 0731, are MIT open weights on Hugging Face.
List pricing roughly 36×–179× cheaper on input and ~89× on output vs Opus 4.8 per million tokens (vendor rates, not audited).
No confirmed date—changelog says "as soon as possible." August rumors unconfirmed.
SWE-bench Verified is stronger signal. Agent scores need third-party Harness reproduction.
DeepSeek's in-house agent framework (files, tools, engineering tasks)—named July 31, not yet public.
V4-Flash-0731 proves post-training and pricing can matter more than another parameter bump—but validating OpenClaw multi-model routing, comparing Harness vs Claude Code, or batch-migrating APIs still benefits from a real macOS graphical session. If your daily driver is Windows or Linux, buying a Mac just to click through agent permissions is heavy; SSH alone won't cover system dialogs. Rent a VNCMac remote Mac by the hour, wire up DeepSeek V4 in VNC, run agent workloads in isolation, and stop when the project ends. See Mac plans and our SSH-VNC guide; also read our V4 GA overview and Kimi K3 open weights pieces.
Sources: DeepSeek API docs/changelog, technical report, Artificial Analysis (via financial media), 21st Century Business Herald. Verify pricing and release status before publishing—figures as of August 5, 2026.