DeepSeek models

DeepSeek-V4-Pro, DeepSeek-V4-Flash, and DeepSeek-V4.1-Flash models now available on Amazon SageMaker JumpStartGA

AWS · DeepSeek-V4-Pro · DeepSeek-V4-Flash · DeepSeek V4.1 Flash

Messages and Responses APIs add GLM-5.3 and DeepSeek-V4.1-Flash models

Together AI · DeepSeek V4.1 Flash

Workers AI - Planned model deprecations on Workers AIDeprecation

Cloudflare · deepseek-r1-distill-qwen-32b · DeepSeek-V4-Pro · Due Nov 15, 2026

Ultra-fast Serverless mode in Preview, with DeepSeek V4.1 Flash 1017 at 6-8x speed

Fireworks AI · DeepSeek V4.1 Flash

DeepSeek V4 Flash (0731) removed from the shared Serverless Training pool

Fireworks AI · DeepSeek-V4-Flash

Llama-3.3-70B-Instruct-Turbo and DeepSeek-V4-Flash-0731 deprecated on serverless

Together AI · DeepSeek-V4-Flash

Workers AI - DeepSeek V4.1 Flash 1017 now available on Workers AIGA

Cloudflare · DeepSeek V4.1 Flash

DeepSeek V4 Flash (0731) to be removed from Serverless Training on November 2, 2026

Fireworks AI · DeepSeek-V4-Flash

DeepSeek V4.1 Flash 1017 available for LoRA on Serverless Training, 262K context

Fireworks AI · DeepSeek V4.1 Flash

DeepSeek-V4.1-Flash-1017 available on dedicated endpoints

Together AI · DeepSeek V4.1 Flash

DeepSeek-V4.1-Flash-1017 on serverless, with a 1M-token context window

Together AI · DeepSeek V4.1 Flash

DeepSeek-V4.1-Flash-1017 released, with pricing unchanged

DeepSeek · DeepSeek V4.1 Flash

Priority PayGo generally available for Gemini Nano Banana 2.1; DeepSeek-V4.1-Flash in previewGA

Google Cloud · DeepSeek V4.1 Flash

GLM-5.3, GLM-5.3-Flash and DeepSeek-V4.1-Flash on provisioned throughput

Together AI · DeepSeek V4.1 Flash

DeepSeek V4.1 Flash serverless pricing changes: output rises to $1.20 per 1M tokens

Fireworks AI · DeepSeek V4.1 Flash

DeepSeek V4.1 Flash available for LoRA training with up to 262K context

Fireworks AI · DeepSeek V4.1 Flash

Serverless deprecation takes effect for DeepSeek V4, Kimi K2 and GLM 5.2 models

Fireworks AI · DeepSeek-V4-Pro · DeepSeek-V4-Flash

Fireworks serverless decommissions DeepSeek V4, GLM 5.2 and Kimi K2 models on Sept 25

Fireworks AI · DeepSeek-V4-Flash · DeepSeek-V4-Pro

DeepSeek-V4.1-Flash on serverless with 1M context, $0.30 input per 1M tokens

Together AI · DeepSeek V4.1 Flash

DeepSeek-V4.1-Flash released with native multimodal support and lower API prices

DeepSeek · DeepSeek V4.1 Flash · DeepSeek-V4-Flash · DeepSeek-V4-Pro

DeepSeek V4 Flash 0731, Qwen 3.8 27B and Muse Glimmer 30B added for LoRA training

Fireworks AI · DeepSeek-V4-Flash

MiniMax M2.7, GPT OSS 20B and three more models deprecated from serverless

Fireworks AI · DeepSeek-V4-Pro

Four models deprecated on serverless, including DeepSeek-V4-Pro and Kimi-K2.7-Code

Together AI · DeepSeek-V4-Pro

AI Search - New Workers AI text generation models in AI Search

Cloudflare · DeepSeek-V4-Flash · DeepSeek-V4-Pro

DeepSeek-V4-Flash-Vision-Exp experimental vision model available on the API

DeepSeek · DeepSeek-V4-Flash-Vision-Exp

DeepSeek-V4-Pro-0813 on serverless with 1,048,576 context, $1.32 input / $3.96 output

Together AI · DeepSeek-V4-Pro

DeepSeek V4 Flash deprecated from serverless, migrate to DeepSeek V4 Flash (0731)

Fireworks AI · DeepSeek-V4-Flash

Workers AI - DeepSeek V4 Flash and Pro now available on Workers AIGA

Cloudflare · DeepSeek-V4-Pro · DeepSeek-V4-Flash

DeepSeek-V4-Pro generally available, with Responses API and peak/off-peak pricing

DeepSeek · DeepSeek-V4-Pro

DeepSeek-V4-Flash-0731 available on dedicated endpoints

Together AI · DeepSeek-V4-Flash

DeepSeek-V4-Flash-0731 now available for fine-tuning

Together AI · DeepSeek-V4-Flash

LoRA fine-tuning on DeepSeek-V3.1 can target MoE expert layers

Together AI · DeepSeek-V3.1

DeepSeek-V4-Flash now available for fine-tuning

Together AI · DeepSeek-V4-Flash

DeepSeek-V4-Flash-0731 on serverless with 1M context, $0.14 input / $0.28 output

Together AI · DeepSeek-V4-Flash

DeepSeek-V4-Flash API in public beta, with native Responses API support

DeepSeek · DeepSeek-V4-Flash

Many Qwen, Kimi, GLM, DeepSeek and Llama models deprecated for fine-tuning

Together AI · DeepSeek-R1

DeepSeek-V4-Flash available for deployment on dedicated endpoints

Together AI · DeepSeek-V4-Flash

Cached input pricing for GLM-5.1 and Qwen3.5-397B-A17B, DeepSeek-V4-Pro price cut

Together AI · DeepSeek-V4-Pro

Eight legacy serverless models deprecated, decommissioned May 14, 2026

Fireworks AI · DeepSeek-V3.1 · DeepSeek V3.2

DeepSeek-R1, DeepSeek-V3.1 and Qwen3-Coder-Next-FP8 deprecated on serverless

Together AI · DeepSeek-R1 · DeepSeek-V3.1

DeepSeek-V4-Pro and V4-Flash added; deepseek-chat and deepseek-reasoner end 2026-07-24

DeepSeek · DeepSeek-V4-Pro · DeepSeek-V4-Flash

DeepSeek-V4-Pro added to serverless with 512,000-token context, $2.10 input per 1M

Together AI · DeepSeek-V4-Pro

Amazon SageMaker AI now supports serverless reinforcement fine-tuning for 12 additional models

AWS · DeepSeek-R1-Distill-Llama-70B

Amazon Bedrock adds support for six fully-managed open weights models

AWS · DeepSeek V3.2

DeepSeek-R1-0528-tput deprecated and no longer available

Together AI · DeepSeek-R1-0528-tput

DeepSeek OCR, MiniMax M2.1, and Qwen3-VL-8B-Instruct models are now available on SageMaker JumpStartGA

AWS · DeepSeek OCR

DeepSeek-V3-0324 automatically redirected to DeepSeek-V3.1

Together AI · DeepSeek-V3 · DeepSeek-V3.1

DeepSeek-R1-Distill-Llama-70B, Llama 3 70B Turbo and two other models deprecated

Together AI · DeepSeek-R1-Distill-Llama-70B

Kimi-K2, DeepSeek-V3 and DeepSeek-R1 now redirect to upgraded versions

Together AI · DeepSeek-V3 · DeepSeek-R1

DeepSeek V3.2 on serverless, plus cached and uncached token pricing in Model Library

Fireworks AI · DeepSeek V3.2

DeepSeek-V3.2-Speciale available on temporary endpoint until Dec 15, 2025

DeepSeek · DeepSeek-V3.2-Speciale

deepseek-chat and deepseek-reasoner upgraded to DeepSeek-V3.2

DeepSeek · DeepSeek V3.2

deepseek-chat and deepseek-reasoner upgraded to DeepSeek-V3.2-Exp

DeepSeek · DeepSeek-V3.2-Exp

deepseek-chat and deepseek-reasoner upgraded to DeepSeek-V3.1-Terminus

DeepSeek · DeepSeek-V3.1-Terminus

DeepSeek-V3.1 model now available fully managed in Amazon BedrockGA

AWS · DeepSeek-V3.1

Fine-tuning context lengths increased for DeepSeek-R1-Distill, Gemma 3, Qwen and Llama models

Together AI · DeepSeek-R1-Distill-Llama-70B · DeepSeek-R1-Distill-Qwen-14B

Fine-tuning adds gpt-oss-120b, DeepSeek-V3.1 and Llama 4 models

Together AI · DeepSeek-V3.1

DeepSeek-V3.1 added, with fast and thinking modes

Together AI · DeepSeek-V3.1

deepseek-chat and deepseek-reasoner upgraded to DeepSeek-V3.1 with hybrid reasoning

DeepSeek · DeepSeek-V3.1

DeepSeek-R1-0528 serverless inference faster on NVIDIA Blackwell, same pricing

Together AI · DeepSeek-R1

deepseek-reasoner upgraded to DeepSeek-R1-0528, with fewer hallucinations

DeepSeek · DeepSeek-R1

deepseek-chat upgraded to DeepSeek-V3-0324, with higher benchmark scores

DeepSeek · DeepSeek-V3

DeepSeek-R1 is available fully-managed in Amazon BedrockGA

AWS · DeepSeek-R1

deepseek-reasoner now serves the new DeepSeek-R1 model

DeepSeek · DeepSeek-R1