DeepSeek-V4-Pro
Shutting down. DeepSeek-V4-Pro shuts down on Azure AI Foundry on Feb 20, 2028. DeepSeek names DeepSeek-V4.1-Flash as the replacement.
16 mo- Next shutdown
- 494 daysFeb 20, 2028 · Azure AI Foundry
- Arena Text
- 1,458#32 of 128 · Arena
- GPQA Diamond
- 91.7%#22 of 108 · Epoch AI
- List price · $ per 1M
- $1.74 in $3.48 outAzure AI Foundry · cached input $0.145
- Context window
- 1Mtokens
Dates on every platform
| Platform | Model ID and status | Deprecated | Shuts down | Replacement |
|---|---|---|---|---|
| Azure AI FoundryMicrosoft: Model retirement schedule - Microsoft Foundry (opens in a new tab) · checked Oct 14, 2026 | DeepSeek-V4-Pro:2026-04-23Active | – | 16 mo | – |
| Together AITogether AI: Changelog (opens in a new tab) · checked Oct 14, 2026 | deepseek-ai/DeepSeek-V4-ProShut down | – | Shut down | – |
| Together AITogether AI: Changelog (opens in a new tab) · checked Oct 14, 2026 | deepseek-ai/DeepSeek-V4-Pro-0813Active | – | None set | – |
| Fireworks AIFireworks AI: DeepSeek-V4-Pro model page (opens in a new tab) · checked Oct 14, 2026 | accounts/fireworks/models/deepseek-v4-proShut down | – | Shut down | – |
| Fireworks AIFireworks AI: Changelog (opens in a new tab) · checked Oct 14, 2026 | accounts/fireworks/models/deepseek-v4-pro-0813Shut down | Shut down | DeepSeek-V4.1-Flash |
Together names deepseek-ai/DeepSeek-V4-Pro-0813 as the model ID to use instead.
Fireworks named DeepSeek V4 Pro (0813), a later snapshot of the same model, as the migration target. That snapshot left serverless on 25 September 2026.
Removed from serverless only; dedicated deployments of this model are unaffected.
Removed from serverless only; dedicated deployments of this model are unaffected.
What breaks when you move off DeepSeek-V4-Pro
- Availability
- Availability
- Availability
Intelligence
| Benchmark | Publisher | Score | Rank | Published |
|---|---|---|---|---|
| Arena Text | Arena | 1,458 ± 2 | #32 of 128 | Oct 8, 2026 |
| GPQA Diamond | Epoch AI | 91.7% ± 1.5% | #22 of 108 | Aug 18, 2026 |
| Arena WebDev | Arena | 1,447 ± 3 | #37 of 69 | Oct 8, 2026 |
| FrontierMath Tier 4 | Epoch AI | 26.8% ± 7.0% | #31 of 58 | Aug 19, 2026 |
| FrontierMath Tiers 1-3 | Epoch AI | 64.6% ± 2.8% | #31 of 71 | Aug 19, 2026 |
| OTIS Mock AIME 2024-2025 | Epoch AI | 98.6% ± 1.0% | #16 of 105 | Aug 18, 2026 |
| SimpleQA Verified | Epoch AI | 52.9% ± 1.6% | #20 of 68 | Aug 27, 2026 |
| SWE-bench Verified | Epoch AI | 77.6% ± 1.9% | #5 of 28 | Jun 18, 2026 |
Arena (arena.ai), https://arena.ai/leaderboard
Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource].
List prices
| Platform | Input | Cached | Output | In effect | Source |
|---|---|---|---|---|---|
Azure AI FoundryDeepSeek-V4-Pro:2026-04-23 | $1.74 | $0.145 | $3.48 | Aug 1, 2026Current · Global Standard rate from the US East 2 price list. | Microsoft: Azure Retail Prices API (Foundry Models, US East 2) (opens in a new tab) |
Together AIdeepseek-ai/DeepSeek-V4-Pro-0813 | $1.32 | $0.13 | $3.96 | Aug 17, 2026Current | Together AI: Available models (serverless) (opens in a new tab) |
Change history
ShutdownFireworks removes older DeepSeek, GLM, Muse and Kimi models from serverlessFireworks AI1 break
DeepSeek V4 Pro (0813), DeepSeek V4 Flash (0731), DeepSeek V4 Flash Vision Exp, Muse Glimmer 30B, Kimi K2.6, Kimi K2.7 Code and GLM 5.2 left serverless, including their Fast and US-only endpoints. Fireworks named DeepSeek V4.1 Flash, NVIDIA Nemotron 3.5 Lightning 30B A3B, GLM 5.3 and Kimi K3 as migration targets.
- Platform
- Fireworks AI
- Models
- DeepSeek-V4-Pro, DeepSeek-V4-Flash, DeepSeek-V4-Flash-Vision-Exp, Muse Glimmer 30B, Kimi K2.6, Kimi K2.7 Code, GLM-5.2
- Announced
- AvailabilityServerless requests to these model IDs stop working after the date. Dedicated deployments are unaffected.
ShutdownFireworks removes MiniMax M2.7, GPT OSS 20B, DeepSeek V4 Pro and fast Kimi variants from serverlessFireworks AI1 break
MiniMax M2.7, GPT OSS 20B and the original DeepSeek V4 Pro left serverless, along with the Turbo and Fast variants of Kimi K2.6 and the Fast variant of Kimi K2.7 Code. Fireworks named MiniMax M3, GPT OSS 120B, DeepSeek V4 Pro (0813) and the standard Kimi modes as migration targets.
- Platform
- Fireworks AI
- AvailabilityServerless requests to these model IDs stop working after the date. Dedicated deployments are unaffected.
ShutdownDeepSeek V4 Pro preview, Kimi K2.7 Code and Nemotron 3 Ultra leave Together serverlessTogether AI1 break
Together removed deepseek-ai/DeepSeek-V4-Pro (use DeepSeek-V4-Pro-0813 instead), moonshotai/Kimi-K2.7-Code and nvidia/Nemotron-3-ultra-550b-a55b from serverless.
- Platform
- Together AI
- AvailabilityRequests to the removed serverless model IDs fail. DeepSeek V4 Pro and Kimi K2.7 Code stay available on on-demand dedicated endpoints; Nemotron 3 Ultra does not.
Source: Together AI, Changelog (opens in a new tab); Together AI, Deprecations (opens in a new tab)
ReleaseDeepSeek V4 Pro 0813 arrives on Together serverlessTogether AI
deepseek-ai/DeepSeek-V4-Pro-0813 is available on serverless with a context length of 1,048,576 tokens.
- Platform
- Together AI
- Model
- DeepSeek-V4-Pro
Source: Together AI, Changelog (opens in a new tab); Together AI, Available models (serverless) (opens in a new tab)
ShutdownDeepSeek-R1 retired on Azure AI FoundryAzure AI Foundry1 break
Microsoft retired DeepSeek-R1 and lists DeepSeek-V4-Pro as the replacement.
- Platform
- Azure AI Foundry
- Model
- DeepSeek-R1
- Replacement
- DeepSeek-V4-Pro
- AvailabilityInference requests to retired deployments return 410 Gone.
Source: Microsoft, Model retirement schedule - Microsoft Foundry (opens in a new tab); Microsoft, Microsoft Foundry Models lifecycle and support policy (opens in a new tab)