Compare useful capacity, model access, account requirements, rate limits, and post-allowance behavior. Permanent free tiers stay separate from temporary credits and experiments.
“Free” means little without daily tokens, requests, model access, and reset behavior.
Prototype vs production
Experimental endpoints and rotating models are useful, but clearly labeled as unstable.
Official evidence
Every changing claim links to the provider and carries a verification date.
Clear categories
Different kinds of “free” are not interchangeable
Filter verified offers
Verified offers
Showing 9 of 9; commission never affects ranking.
Free Tier
Hetzner
Verified
Free experimental OpenAI-compatible inference
Collected official Hetzner documentation reconfirms free OpenAI-compatible inference while experimental, with Qwen/Qwen3.6-35B-A3B-FP8 and Qwen3.8-27B documented at 262,144-token context with image and text input. Customers need Experiments access and an API token; availability is not guaranteed, production use is discouraged, and the authenticated /v1/models response is definitive for the rotating model list.
Value
No-cost experimental hosted inference; two documented vision models with 262,144-token context
Who qualifies
Hetzner customers with Experiments access and an inference API token; no startup or VC requirement stated
Expiry
Free while experimental; advance email notice promised if status changes; no fixed deadline stated
Smartoolbox does not currently earn a commission from this offer.
Free Tier
OpenRouter
Verified
Free-model router and zero-cost model variants
The zero-price listing adds HeyGen Voice Turbo (dated October 26), Recursive Flint, and NVIDIA Nemotron 3.5 Embed 1B and 8B (dated October 21), all absent from the master. Voice Turbo supports 39 languages and up to 5,000 characters per request and is marked as going away November 30; Flint lists 262K context and the embedding models return 2048- and 4096-dimensional vectors respectively.
Value
$0/M input and $0/M output tokens on four listed hosted models; Free plan lists 50 requests/day
Who qualifies
OpenRouter users with access to the relevant API models, subject to account request caps and model availability; own HeyGen key needed for voice clones from the user's account
Expiry
HeyGen Voice Turbo: going away November 30, 2026; other listed models: not stated
Smartoolbox does not currently earn a commission from this offer.
Credits
Anthropic
Verified
Claude Max and Ultra advertise $100 and $200 monthly API credits
The official pricing excerpt advertises $100 monthly Claude Platform API credits with the $100 Max subscription and $200 monthly API credits with the $200 Ultra subscription. These are paid-plan benefits, not unconditional free credits; the referenced credit footnote is absent from the supplied excerpt.
Value
$100/month API credits on $100 Max; $200/month on $200 Ultra, before applicable tax
Who qualifies
Claude Max or Ultra subscribers, subject to API-credit terms referenced by the pricing page; no startup or VC requirement stated
Expiry
not stated; monthly benefit, with rollover and credit-expiry terms not captured
Smartoolbox does not currently earn a commission from this offer.
Free Tier
Google
Verified
Gemini 4 Carbon launch pricing halves token rates through November 30
Google's collected official pricing lists Gemini 4 Carbon input at $2/M tokens and output at $10/M through November 30, 2026, rising to $4/M and $20/M on December 1. Carbon is not available on the Free tier, and thinking tokens and internal tool instructions are billed.
Value
50% lower input/output rates through November 30; implicit caching $0.10/M versus $0.20/M from December 1
Who qualifies
Paid Gemini API users with access to gemini-4-carbon; not available on the Free tier
Expiry
November 30, 2026; higher listed rates start December 1, 2026
Smartoolbox does not currently earn a commission from this offer.
Free Tier
Nous Research
Verified
Nous Portal reconfirms five zero-priced hosted models on November 9
The authenticated source-pack snapshot reconfirms zero prompt and completion prices for meituan/longcat-2.5:free, reflection/beam-1.0-501b-a23b:free, stealth/crouton-alpha, stepfun/step-3.7-flash:free, and upstage/solar-mini4:free. This is one rotating Portal catalog, unchanged from November 8; numeric quotas and production guarantees are not stated.
Value
Five hosted models at $0 prompt and $0 completion pricing at collection time
Who qualifies
Nous Portal account holders with catalog access; sign in through the official Portal
Smartoolbox does not currently earn a commission from this offer.
Free Tier
Nous Research
Verified
Nous Portal free catalog contracts from five to three hosted models
The authenticated collection lists meituan/longcat-2.5:free, stepfun/step-3.7-flash:free, and upstage/solar-mini4:free at zero input and output prices. Compared with the November 9 master snapshot, Reflection Beam and stealth/crouton-alpha are no longer in this zero-priced catalog; availability can rotate.
Value
Three hosted models with $0 prompt and completion pricing at collection time
Who qualifies
Nous Portal account holders with model-catalog access; numeric quotas and production guarantees not stated in the live snapshot
Smartoolbox does not currently earn a commission from this offer.
Free Tier
OpenRouter
Verified
OpenRouter lists Crouton Alpha with zero-cost tokens and 1.05M context
OpenRouter's collected official free-model collection lists stealth/crouton-alpha at $0/M input and output tokens with a 1.05M-token context window for coding, agentic, and multimodal workflows. The anonymous third-party provider may retain prompts and completions but says they are not used for training; free capacity and account request caps apply.
Value
$0/M input and output tokens; 1.05M-token context
Who qualifies
OpenRouter users with an API key and access to the free model; no startup or VC requirement stated
Expiry
not stated; stealth preview availability may change
Smartoolbox does not currently earn a commission from this offer.
Free Tier
Together AI
Verified
Together AI reconfirms zero-cost Ternary Bonsai 27B inference
The collected official Serverless Inference pricing table still lists Ternary Bonsai 27B at zero input and output token cost, refreshing an offer last seen August 28. Quotas, access restrictions beyond platform access, and an end date are not stated in the supplied excerpt.
Value
$0/M input tokens and $0/M output tokens
Who qualifies
Together AI users with access to its Serverless Inference API
The collected official pricing table advertises DeepSeek-V4.1-Flash-1017 at 51% off, with $0.098/M input, $0.0029/M cached input, and $0.294/M output. This is a different discounted model from the already-known Flash Vision Experimental offer; the table also marks DeepSeek-V4.1-Flash at 30% off.
Smartoolbox does not currently earn a commission from this offer.
Need a durable comparison?
The maintained guide compares model access, useful capacity, card requirements, data treatment, regions, and what happens when the free allowance ends.
