Pollinations AI

LLM 4.4/10 Jul 4, 2026

Open-source Berlin-based gen-AI gateway for free text, image, audio, and video generation with no sign-up required and OpenAI-compatible endpoints.

Requests / min: 4
Card required: No

Pollinations AI scores 4.4/10 for free value in LLM. The main constraint: Anonymous (no sign-up) access is rate-limited to roughly one request every 15 seconds per IP on basic models across text, image, audio, and video; free registration (Seed tier) raises this on standard models.

Category LLM Kind Gen-AI API gateway Caveats No token-based quota — access is governed by per-IP/per-request rate limits, not a monthly token allowance. Free-tier images may carry a Pollinations watermark (removable only with an account). The newer "Pollen" credit system ($1 ≈ 1 Pollen) and Flower/Nectar tiers gate advanced models and unmetered usage behind paid plans, and publishable client-side keys are not yet recommended for production. Card required No Eligibility anyone

Free value score

4.4 / 10

Show the math

Model Quality 2.7/10

value = 15 artificial_analysis_score

weighted score = 0.5 * 2.7/10

Request Rate 2.2/10

value = 4 req/min

weighted score = 0.25 * 2.2/10

Access 10/10

value = 10/10

weighted score = 0.25 * 10/10

Key metrics

GPT-OSS 20B (openai-fast) 15 Best model

4 R/min Requests/min

Yes Commercial use

Models

1 free · 106+ total

Limits may be stale — last re-verified outside this section's freshness window.

Model	Family	AA	Access
GPT-OSS 20B (openai-fast)	GPT	14.9	free
claude-large (Claude Opus 4.8)	Claude	55.7	paid
openai-large (GPT-5.5)	GPT	54.8	paid
gpt-5.4 (GPT-5.4)	GPT	51.4	paid
glm (Z.ai GLM-5.2)	GLM	51.1	paid
gemini (Gemini 3.5 Flash)	Gemini	50.2	paid
gemini-large (Gemini 3.1 Pro)	Gemini	46.5	paid
minimax (MiniMax M3)	MiniMax	44.4	paid
deepseek-pro (DeepSeek V4 Pro)	DeepSeek	44.3	paid
claude-opus-4.7 (Claude Opus 4.7)	Claude	42.7	paid
kimi-code (Kimi K2.7 Code)	Kimi	42	paid
deepseek (DeepSeek V4 Flash/Lite)	DeepSeek	40.3	paid
gpt-5.4-mini (GPT-5.4 Mini)	GPT	40	paid
qwen-large (Qwen3.7 Plus)	Qwen	39	paid
openai (GPT-5.4 Nano)	GPT	38.2	paid
claude-opus-4.6 (Claude Opus 4.6)	Claude	37.8	paid
grok-large (Grok 4.3)	Grok	37.6	paid
claude (Claude Sonnet 4.6)	Claude	35.9	paid
glm-step / step-flash (StepFun Step 3.7 Flash)	StepFun	29.7	paid
gemini-3-flash (Gemini 3 Flash)	Gemini	27.4	paid
gemini-flash-lite-3.1 (Gemini 3.1 Flash Lite)	Gemini	25	paid
mistral (Mistral Small 4)	Mistral	20.8	paid
mistral-large (Mistral Large 3)	Mistral	16.2	paid
llama-maverick (Llama 4 Maverick)	Llama	14.3	paid
gemini-fast (Gemini 2.5 Flash Lite)	Gemini	14.1	paid
gemini-search (Gemini 2.5 Flash Lite Search)	Gemini	14.1	paid
llama-scout (Llama 4 Scout)	Llama	10	paid
claude-fast (Claude Haiku 4.5)	Claude	—	paid
flux (Flux Schnell)	Flux (Black Forest Labs)	—	paid
grok (Grok 4.20 Non-Reasoning)	Grok	—	paid
grok-4-20-reasoning (Grok 4.20 Reasoning)	Grok	—	paid
kimi (Moonshot Kimi K2.6)	Kimi	—	paid
llama (Llama 3.3 70B)	Llama	—	paid
nanobanana-pro (NanoBanana Pro)	Gemini/Google	—	paid
nova (Amazon Nova 2 Lite)	Nova	—	paid
perplexity (Perplexity Sonar Pro)	Perplexity	—	paid
qwen-coder (Qwen3 Coder 30B)	Qwen	—	paid
qwen-coder-large (Qwen3 Coder Next)	Qwen	—	paid
qwen-vision (Qwen3 VL 30B Instruct)	Qwen	—	paid
qwen-vision-pro (Qwen3 VL 235B Thinking)	Qwen	—	paid

Showing first 1 of a larger catalog.

Compare all free models across providers →

What you get

No sign-up required for basic text, image, audio, and video generation

OpenAI-SDK-compatible text endpoint plus an official MCP server for Claude/Cursor/Codex

Broad model roster: GPT/Gemini/Llama/Mistral (text), Flux/gpt-image/Seedream (image), wan-fast (video), 30+ TTS voices

100% open-source (MIT) stack, no credit card needed to start

Gotchas

Watermark on free images

Since March 31, 2025 free-tier images may include a Pollinations watermark; removing it (nologo) requires a registered account.

Sources

"| Anonymous | One request every 15s | Basic models | No signup | Good for testing | Seed registration raises limits on standard models. |"
https://raw.githubusercontent.com/pollinations/pollinations/master/APIDOCS.md
"[{"name":"openai-fast","description":"GPT-OSS 20B Reasoning LLM (OVH)","reasoning":true,"tier":"anonymous",...}]"
https://text.pollinations.ai/models
"i used to have 1 pollen per day and its 1.5 pollen per week now. why? how to raise my pollens? bc my bot is using AI services"
https://github.com/pollinations/pollinations/issues/8925
"gpt-oss-20b (20.9B total parameters, 3.6B active parameters): Intelligence Index score of 48"
https://artificialanalysis.ai/models/comparisons/gpt-oss-20b-vs-o1-preview

llm image-generation text-to-speech video-generation open-source openai-compatible mcp no-signup multimodal free-tier

Open service Source ← All LLM

Alternatives

Same category · similar quota

Sarvam AISarvam AI · finite_credit · Indic AI API platform

4.5

India's sovereign full-stack AI platform offering speech-to-text, text-to-speech, translation, and LLM/chat APIs across 22 Indian languages.

$1.20 one-time*11.9 AA40M tok

May be outdatedChecked 27d ago

One-timeDead

SiliconFlowSiliconFlow · freemium · Hosted open-weight inference (CN)

4.7

China-based inference platform offering three permanently free open-weight models alongside 200+ paid models.

15 AA30/min

Checked 10d ago

OpenRouter CreditsOpenRouter Credits · paid · Paid plan

4.8

Pre-paid credits routed to 300+ models at provider list prices; credits never expire; Kimi K2 Thinking is $0.60/$2.50 per 1M on OpenRouter; 1M free BYOK requests/mo then 5% BYOK fee

930.2K/$19.4 AA930.2K/$

May be outdatedChecked 26d ago

Paid onlyEmbeddings