AI & machine learning

$0 Starter: $10 model credits/month, 1 GB processed data, 10k scores, 14-day retention and unlimited users/projects. No card.

Free inference tier with access to all Cerebras-powered models, lower rate limits and community support.

10,000 Neurons/day free across eligible models; usage above that is $0.011/1k Neurons on Workers Paid, and some heavy models require Paid.

Cloud Free: up to 10 members, 25k spans/month and 60-day retention; open-source self-hosting remains free.

No free tier or trial for Model APIs; fal uses prepaid credits. New accounts start at 2 concurrent requests and scale to 40 as paid invoices accumulate.

New accounts get $1 in free credits. Without a payment method or active credits, the limit is 10 requests per minute; with credits, 6,000 per minute.

Free with no subscription, API key or card; text ads fund usage and accounts receive 100 Freebucks/day, resetting at midnight Pacific.

AI Studio is free to use; selected Gemini API models include a free tier with model-specific rate limits.

Free API plan with model-specific request and token limits; exact quotas are published in the Groq console docs.

Free Hub account with public/private repos, free CPU Basic or ZeroGPU Spaces and $0.10/month of Inference Providers credits.

Free entry plan is published at $0; Respan combines its multi-model gateway, tracing, monitoring and evaluation tools, with model inference billed separately.

Hobby: $0, 2 users, 50k usage units/month and 30-day data access; open-source self-hosting is also free.

$0 Free plan with $10/month in API credits for experimentation and lower limits; Labs models are free but experimental.

$0 Starter plan with $30 per month in free credits, 3 seats, 100 containers and 10 GPU concurrency. A payment method is required, and extra usage bills by time.

100 notebooks with up to 50 sources each; since Sep 2026 AI interaction quotas are compute-based, refresh every 5 hours and are shown in-app.

Free hosted NIM API endpoints for prototyping through the NVIDIA Developer Program; available models and limits can change.

Free: 25+ free models, 4 free providers, 20 requests/min and 50/day; buying at least $10 in credits raises free-model daily use to 1,000.

API key required. Usage is metered in Pollen credits, earned through Quests or purchased; publishable keys are available for client apps.

Free Forever: 10k recorded logs/month, 3-day logs, 30-day metrics, gateway/fallbacks/retries, simple cache and guardrails.

Some models run free before billing; prediction endpoints allow 600 requests per minute and other endpoints 3,000. No fixed free-run quota; billing required once it ends.

No free tier or trial; a minimum $5 prepaid credit purchase is required. Serverless inference uses dynamic per-model rate limits, and access pauses at a zero credit balance.