TypeTime-limited7Free Tier123New User28Free Only7
AccessAPI57Desktop37CLI22Web87App21
TagCoding29Agents30Work26Chat30Image31Video24Audio23Search16Embedded3Platforms21OCR4Scraping6Document Parsing6TTS8Compute2

Free AI Platforms: Free Tiers, Limits

Free AI Platforms

21 resources
Free: up to 40 requests/min for most models, no per-token billing; personal limit shown in dashboard

Serverless inference APIs accelerated by DGX Cloud

    Some models are time-limited free: Spark-X2.5-4B (native 1M token context), DeepSeek-OCR.

    iFlytek MaaS platform with Spark and 60+ open-source models

      :free models: 20 req/min; 50/day before top-up, 1,000/day after $10 purchased

      Unified API gateway with hundreds of models

        Free to use Free access to numerous AI generators; many AI image, chat, and story generators require no sign-up and have unlimited use.

        Browser AI image generator on Perchance, usable without an account

          User monthly allowance A free monthly allowance per user (exact amount shown in the usage dashboard); upgrade prompts once exhausted

          User-pays model: call GPT/Claude models from frontend scripts

            2,000 free API calls per day, 200 MoLi credits on login (profile completion required)

            Alibaba's open-source model hub for hosting and inference

              Free plan: 10,000 Neurons per day

              Serverless edge inference for Llama, Qwen, GLM, Kimi

                Starter plan: $30 free compute credits/month (no rollover)

                Serverless GPU compute platform built for AI/ML workloads

                  Curated models in the try-for-free collection come with a fixed number of free runs, not free forever (No. The models in this collection are free for a limited number of runs); after the free use limit you must add billing and purchase credit to continue.

                  Serverless cloud GPU inference for hundreds of open models

                    Anonymous access No signup and no API key: call the image.pollinations.ai and text.pollinations.ai endpoints directly for image/text generation; official rate limits are not published.
                    Quest Pollen Earn free Pollen by completing GitHub Quests, trying models, or publishing an app/agent (1 Pollen ≈ $1) to spend on paid models; quests and rewards change dynamically.

                    Text / image / audio generation API, no signup, open source

                      Free tier: $5 usage per month (selected models only; not applicable after purchasing paid credits)

                      Unified AI gateway: hundreds of models behind one endpoint

                        Free shared model APIs: 20 req/min per account (gateway) + per-period spend cap

                        AMD's official GPU cloud for model inference

                          Sandbox free: 200 message credits, 5 apps, 50 knowledge docs, 50MB storage

                          Open-source AI app platform, visual Agent + RAG

                            Free plan Limited messages, web searches and coding sessions on Vibe (web + mobile), basic image generation, and Mistral Studio to test Mistral models

                            Mistral's official API, chat and coding platform

                              3,000 resource points per month (1,000 resource points ≈ ¥1); 1 free environment per account, each renewed for 6 months

                              DB, functions, storage and LLM calls on one resource-point pool

                                Some models free The pricing page marks each model as free or time-limited free (including some sub-9B small models); check the per-model labels for the exact scope.
                                Until New user credit 16 CNY voucher on real-name verification; +16 CNY each for invited friends (until 2026-12-31)

                                AI model aggregation platform with custom inference engine

                                  $200 credits (30 days) + popular services free for 12 months (see official list)

                                  Microsoft Azure account for new users with AI services

                                    1M shared tokens on activation (valid 1 year) + another 1M for Embedding

                                    Tencent Hunyuan LLM API, incl. Hy4 preview

                                      1M tokens per model, valid 90 days (Beijing region only)
                                      Auto-granted after real-name verification on first activation; pay-as-you-go afterwards

                                      Model marketplace with 70+ models including Qwen

                                        U2 Decision free for a limited time Time-limited free: input, cache hits and output are all ¥0 per 1M tokens (no end date announced)

                                        Unisound model platform, U2 Flash and more free for a limited time

                                          New users can claim 10M tokens from the registration promotion.
                                          Some free LLM models are provided, with rate or frequency limits.

                                          LLM API service on China's national supercomputing network