DeepSeek V4 Pro on VM0. Cost-optimised reasoning
DeepSeeks Flaggschiff. Erstklassige SWE-bench-Ergebnisse, 1M-Kontext und Open-Source-Lizenz — das beste Preis-Leistungs-Verhältnis für Code-Agenten.
1M tokens · Text / Code · Prompt cache
DeepSeek V4 Pro ist das Flaggschiff von DeepSeek, veröffentlicht am 24. April 2026. Es bietet Spitzen-Code-Benchmarks (SWE-bench Verified, Terminal-Bench, LiveCodeBench), ein 1M-Token-Kontextfenster und eine Open-Source-Lizenz — alles zu einem aggressiven Preis.
Listenpreis $1,74/$3,48 pro 1M Tokens mit gecachtem Input bei $0,145/1M. Cache Writes sind kostenlos. Auf VM0 bei ×0,3 Credits positioniert, ist es eine der kosteneffizientesten Optionen für Code-schwere Agent-Workflows.
Was ist DeepSeek V4 Pro?
24. April 2026 · Flaggschiff der DeepSeek V4-Familie. Pro-Variante mit maximaler Reasoning-Tiefe.
DeepSeek V4 Pro is the flagship of DeepSeek's V4 generation, released April 24, 2026 under the MIT License. It's an open-weight Mixture-of-Experts model with 1.6T total parameters and 49B active per token, paired with V4 Flash (284B / 13B active) for cost-sensitive work.
Both V4 models share an identical feature set: 1M-token context window, 384K maximum output, three reasoning effort modes (standard, think, think-max), JSON output, tool calls, and FIM completion in non-think mode. The Pro model adds a hybrid attention architecture (Compressed Sparse Attention + Heavily Compressed Attention) for dramatically improved long-context efficiency. 27% of single-token inference FLOPs and 10% of KV cache vs DeepSeek V3.2 at 1M context.
DeepSeek made waves through 2025 by delivering Anthropic-grade reasoning at a fraction of the price. V4 Pro continues that pattern: vendor-reported SWE-bench Verified 80.6% sits within 0.2 points of Claude Opus 4.6, at roughly one-seventh the vendor cost. On VM0 it's exposed via the DeepSeek API-key provider and on VM0 Managed at ×0.3. The same multiplier as Claude Haiku 4.5 but with substantially stronger reasoning behaviour.
Technische Daten auf einen Blick
DeepSeek V4 Pro Benchmarks
Vendor-reported scores from DeepSeek's V4 Pro release. Independent reviews (Geeky Gadgets, Code Arena) place V4 Pro third on Code Arena behind GLM-5.1 and Kimi K2.6. The strongest benchmark claims come from DeepSeek's own materials. Treat directionally rather than as absolute truth.
DeepSeek V4 Pro Preise
Listenpreis des Anbieters, pro 1 Mio. Tokens.
Wie sich DeepSeek V4 Pro in der Praxis verhält
Beobachtetes Verhalten aus produktiven Agent-Durchläufen.
Reasoning
Strongest sub-Sonnet reasoning in our lineup. Holds up on multi-step work where cheaper models start to drift. Vendor-reported MMLU-Pro matches GPT-5.4.
Coding benchmarks
Vendor-reported SWE-bench Verified 80.6% (within 0.2 of Opus 4.6), Terminal-Bench 2.0 67.9% (leads Opus 4.6), LiveCodeBench 93.5%.
Cost efficiency
The standout property. ×0.3 credit cost with reasoning that competes well with Sonnet 4.6 makes V4 Pro the cost-optimisation default. ~7× cheaper than Claude Opus 4.7.
Cache economics
Cache writes are free. Unique among VM0's Built-in models. Stable system prompts and large pasted reference docs cost nothing extra to cache, only the read side bills.
Speed
Around 36 tokens/sec at max effort per Artificial Analysis. Slower than Haiku, slightly slower than Opus 4.6.
Beste Agent-Aufgaben für DeepSeek V4 Pro
The PR-review agent that runs on every commit
Sonnet-tier accuracy at roughly one-third of Sonnet's vendor cost is what makes "review every commit, not just the big PRs" actually viable. V4 Pro reads the diff, the related files, and the linked issue, then writes a structured comment — and the per-call price is low enough that running it as a CI step on every push doesn't show up as a noticeable line item.
The scheduled summariser that runs every night
Pulls yesterday's customer conversations, support tickets, or sales calls and writes a digest. The system prompt and tool schema don't change between runs, and DeepSeek doesn't bill cache writes — so the long fixed prefix is paid for once and cached reads cost a fraction of normal input. This is where V4 Pro's pricing model genuinely changes what's affordable.
The whole-repo code agent that costs less than Opus
1M-token context with hybrid attention (Compressed Sparse Attention plus Heavily Compressed Attention) means a mid-sized codebase fits in one prompt and inference cost stays manageable as the window fills up. For cross-file refactors and architecture-level reviews, this is where you get the Opus-style "see everything at once" workflow without the Opus-style invoice.
Wann du DeepSeek V4 Pro überspringen solltest
Skip V4 Pro on the hardest tool-routing edge cases where Sonnet 4.6 still leads, and on bulk single-shot work where reasoning isn't required and V4 Flash is roughly 12× cheaper.
DeepSeek V4 Pro vs andere Modelle
DeepSeek V4 Pro vs DeepSeek V4 Flash
Same vendor, different positioning. V4 Pro (×0.3) gives you reasoning; V4 Flash (×0.02) gives you the cheapest possible single-shot model. Vendor-reported SWE-bench Verified shows Flash within 1.6 points of Pro (79.0 vs 80.6). But Pro pulls ahead on Terminal-Bench (67.9 vs 56.9) on multi-step tool use.
DeepSeek V4 Pro vs Claude Sonnet 4.6
Sonnet 4.6 (×1) wins on tool-routing edge cases and English-language reasoning. V4 Pro (×0.3) wins on cost and is competitive on coding benchmarks (vendor-reported). Worth A/B-testing on a real agent before committing.
DeepSeek V4 Pro vs Kimi K2.6
Same multiplier (×0.3). Kimi has stronger long-context recall and a higher Intelligence Index (54 vs 52); V4 Pro has the better cache economics (free writes) and a 1M context window vs Kimi's 256K. Pick by which property matters more.
Fazit: Solltest du DeepSeek V4 Pro nutzen?
DeepSeek V4 Pro ist die erste Wahl für Code-intensive Agenten, wenn die Gesamtbetriebskosten im Vordergrund stehen. Seine Open-Source-Lizenz und führenden Code-Benchmarks machen es zum stärksten Kosteneffizienz-Kandidaten auf VM0.
Häufig gestellte Fragen
When was DeepSeek V4 Pro released?
DeepSeek released V4 Pro and V4 Flash together on April 24, 2026 under the MIT License with open weights.
Why are cache writes free?
DeepSeek doesn't bill the cache-write portion. Only cache reads bill, at $0.145 per 1M tokens. Stable system prompts and large reference contexts cost nothing extra to cache.
What's V4 Pro's context window?
1 million tokens with up to 384K tokens of output. The hybrid attention architecture makes the full window usable at much lower inference cost than V3.2.
How does V4 Pro compare to Claude Opus 4.6?
Vendor-reported SWE-bench Verified is within 0.2 points (80.6 vs 80.8). Terminal-Bench 2.0 favours V4 Pro (67.9 vs 65.4). Opus 4.6 leads on HLE (40.0 vs 37.7) and HMMT 2026 math (96.2 vs 95.2). At ~7× lower vendor cost, V4 Pro is the right call when reasoning quality is the bar but cost matters.
Is V4 Pro open-source?
Yes. Weights are published under the MIT License. The hosted DeepSeek API is the production path for VM0.
Alternativen
DeepSeek V4 Pro auf VM0 nutzen
Zwei Wege, um DeepSeek V4 Pro auf VM0 zu nutzen
VM0 unterstützt DeepSeek V4 Pro als Built-in-Modell, das in VM0-Credits abgerechnet wird, sowie über Bring-your-own mit einem DeepSeek API key. Der Built-in-Weg nutzt VM0 Managed Routing und den unten erklärten Credit-Multiplikator; der Bring-your-own-Weg rechnet direkt mit dem Upstream-Anbieter ab und überspringt die VM0-Credit-Umrechnung.
VM0s Empfehlung
VM0 positioniert DeepSeek V4 Pro als kostensparende Option statt als Core-Agent-Modell. Nutze es zur Optimierung der Stückkosten bei Nicht-Kernarbeit wie Massenklassifikation, Vorfiltern, latenzkritischen Kurzantworten oder fest zugewiesenen Legacy-Agents, während Claude Opus 4.7, Claude Opus 4.6 oder Claude Sonnet 4.6 die entscheidenden Schritte übernehmen.
Credits und der ×0.3-Multiplikator
Jedes Built-in-Modell auf VM0 wird als Vielfaches von Claude Sonnet 4.6 bepreist, das die ×1-Credit-Basislinie bildet. DeepSeek V4 Pro wird mit ×0.3 Credits abgerechnet. Der Multiplikator erscheint auf deiner VM0-Rechnung; der Anbieter-Listenpreis in der obigen Preistabelle ist das, was der Upstream-Anbieter berechnet, bevor VM0 ihn in Credits umrechnet.
DeepSeek V4 Pro wird mit ×0.3 abgerechnet, d.h. ein Schritt kostet hier nur das 0.3-fache der Credits eines äquivalenten Schritts mit Sonnet 4.6 (der ×1-Basislinie). Damit liegt es deutlich unter der Credit-Basislinie und ist die natürliche Wahl für volumenstarke Hintergrundarbeit, bei der Kosten pro Schritt wichtiger sind als höchste Reasoning-Qualität.
Verfügbar auf VM0 seit April 24, 2026.