Inference
Model APIs that answer a completion request, keyless or behind a free key.
Best per needInference APIs with no card and no phone
Key
- ★ qualifies
- usable as stated; the read path was probed
- caveat
- usable, with a named limit or an unprobed read path
- trial
- a time-boxed or one-time trial, not a recurring allowance
- paid
- no free allowance; listed for its price
- fails: identity
- the identity required is not reachable without it
- fails: egress
- data cannot be read back out at a usable rate or price
- fails: quota
- the free allowance is too small to matter
- dead
- the offer or the service no longer exists
- self-hosted, last resort
- software or a node that the user installs, runs and maintains. Self-hosted options are listed last, because this site is for users who have no hardware.
- free tier, no free tier
- the record states a free allowance, or states that it has none
- identity email
- what a signup asks for. A question mark after it: a placeholder, not read from a source
- no card, card, card?
- a payment card is not asked, is asked, or the record does not say
- limit
- the limit that runs out first, for example request-rate or volume-quota
- evidence
- how the record is backed, strongest first: measured, documented, reported, inferred, unverified
- fresh until, stale since
- the check is inside its window, or past it, on today's date
- inherited, checked live
- inherited: the record was carried from an earlier source, and its check date is an upper bound. Checked live: the record was checked against the service. The chip filters on it
- verdict
- a judgement of the whole offer for the need of its topic. The data holds no number as a cut-off. Read the free allowance and the limits of the record, and decide for your own need
- duplicate, disagrees
- another record holds the same offer, or a sibling record states another value
Usable 40
- ★ qualifiesOVHcloud AI Endpoints (anonymous tier)
No key required; 22 model ids on /v1/models, each carrying a pricing object in USD per token.
free tieridentity anonymousno cardlimit request-rateevidence measuredstale on 2026-11-28
- caveatBlockRun
7 of 117 model ids on /v1/models carry an input price of 0; the rest carry list prices spanning openai, google, anthropic, xai, zai, bytedance, deepseek...
free tieridentity anonymousno cardlimit request-rateevidence measuredstale on 2026-12-05
- caveatfree.ai
513 model entries in /v1/models, 45 of them marked self_hosted true; every id probed answers from one backend, Qwen/Qwen3-30B-A3B-Instruct
free tieridentity anonymousno cardevidence measuredstale on 2026-11-28
- caveatKilo Gateway
14 of the 401 ids in the catalogue end in ':free' and require no key; three more are free without the suffix (reflection/beam-free, kilo-auto/free...
free tieridentity anonymousno cardlimit request-rateevidence measuredstale on 2026-12-05
- caveatLLM7
No key required for the ids probed; /v1/models lists 68 ids, none with a zero price
free tieridentity anonymousno cardlimit request-rateevidence measuredstale on 2026-11-28
- caveatPollinations
The Classic API lists one model, openai-fast, described as 'GPT-OSS 20B Reasoning LLM (OVH)', with aliases openai, gpt-oss, gpt-oss-20b and ovh-reasoning
free tieridentity anonymousno cardlimit request-rateevidence measuredstale on 2026-11-28
- caveatAI Horde (text generation)
Text generation on volunteer GPUs with no account: the README gives API key 0000000000 for anonymous use and 'No need to register'.
identity anonymousno cardlimit request-rateevidence documentedstale on 2026-11-28
- caveatAion Labs
A Free tier at $0 with a daily credit allowance; the Free rate tier is 20,000 tokens per minute and a 20,000 daily token limit, no card
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-05
- caveatApi.Airforce
Free plan, $0.00/month: 1 request per minute, 1 000 requests per day, free models only
free tieridentity anonymousno cardlimit request-rateevidence documentedstale on 2026-11-28
- caveatBazaarLink
5 enabled free models at the time read; accounts without credit get 10 requests per minute and 60 weighted units a day, accounts with credit 20 per minute...
free tieridentity emailcardlimit volume-quotaevidence documentedstale on 2026-12-05
- caveatBeaverNet
Works without a key, with limits; a free key gives more. 4 models available; the listing carries a per-model allow_anonymous flag
free tieridentity anonymousno cardlimit request-rateevidence documentedstale on 2026-11-28
- caveatCartesia (Free)
Free plan at USD 0 a month with 20K credits per month plus USD 1 of prepaid voice-agent usage per month.
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-04
- caveatCloudflare Workers AI (Inference)
10,000 Neurons per day at no charge on both the Free and the Paid Workers plan.
free tieridentity emailno cardlimit request-rateevidence documentedstale on 2026-11-28
- caveatDG AI API (dg-ai.scriptsnsenses.workers.dev)
Free public beta with no key required: 18 models live, auth is optional and a placeholder of dashes is accepted; the anonymous tier is 8 requests per 10 s
free tieridentity anonymous?card?limit request-rateevidence documentedstale on 2026-11-28
- caveatElevenLabs (Free)
Free plan at USD 0 with 10,000 credits per month, no credit card required.
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-04
- caveatflm (flm.stonech.at)
No account and no API key.
free tieridentity anonymousno cardlimit request-rateevidence documentedstale on 2026-11-28
- caveatGoogle Cloud Translation
The first 500,000 characters per month of language detection and text translation with the NMT model are free, applied as a $10 credit every month; the...
free tieridentity cardlimit volume-quotaevidence documentedstale on 2026-12-04
- caveatGoogle Gemini API (Inference)
Gemini-3.8-flash is 'Free of charge' for input and output on the Free Tier
free tieridentity emailno cardlimit request-rateevidence documentedstale on 2026-12-01
- caveatGroq
A Free Plan per model, printed as RPM / RPD / TPM / TPD in the page's own units: openai/gpt-oss-120b and openai/gpt-oss-20b at 30 / 1K / 8K / 200K...
free tieridentity emailno cardlimit request-rateevidence documentedstale on 2026-12-01
- caveatHugging Face public Gradio chat Spaces (anonymous API)
ZeroGPU Spaces give an unauthenticated caller 2 minutes of GPU quota a day at low queue priority (5 minutes for a free account); a public Gradio chat Space...
free tieridentity anonymousno cardlimit volume-quotaevidence documentedstale on 2026-11-28
- caveatHugging Face ZeroGPU Spaces (public Spaces called through the Gradio client)
ZeroGPU Spaces are available to use for free to all users: 2 minutes of GPU a day unauthenticated and 5 minutes a day on a free account, reset exactly 24...
free tieridentity anonymousno cardlimit volume-quotaevidence documentedstale on 2026-11-28
- caveatIBM watsonx.ai (Free / Lite plan)
Free plan (Toolbox playground): foundation-model tokens up to 300,000 per month; the embeddings and reranking APIs are listed as available on the free Lite plan
free tieridentity cardlimit volume-quotaevidence documentedstale on 2026-12-04
- caveatIteraCompute
Free Qwen3.8-27B: 30M tokens a month, 128K context, 160,000 tokens per minute, no credit card; the promotion ends at 00:00 Pacific Time on November 24, 2026
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-05
- caveatJarvisClaw free route (auto/free)
Requests with model auto/free and no credential are not billed and are rate limited per IP address in gateway calls per minute; no number is published
free tieridentity anonymousno cardlimit request-rateevidence documentedstale on 2026-11-28
- caveatLlamaParse
Free plan, USD 0 a month: 10,000 credits a month, recurring (credits reset at the start of the next billing cycle).
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-04
- caveatMistral La Plateforme (Inference)
Free mode is the default plan for new accounts: API access is enabled by default with no credit card required, within usage and rate limits. unverified: no...
free tieridentity emailno cardlimit request-rateevidence documentedstale on 2026-12-01
- caveatNetlify AI Gateway
Available on the Free plan: AI Gateway usage draws on the plan's credit pool, a 300 credit limit on Free, at 180 credits per USD 1 of model usage
free tieridentity emailcard?limit volume-quotaevidence documentedstale on 2026-11-28
- caveatNovita LLM API
Apodex/apodex-1.1-mini (256K context) is priced at $0 per million input tokens and $0 per million output tokens on the serverless LLM API; no token quota is...
free tieridentity emailno cardlimit request-rateevidence documentedstale on 2026-12-04
- caveatOpenRouter :free model ids
Free-model request caps for model IDs ending in :free, selected by all-time credits purchased: 20 requests per minute with 50 requests per day below 10...
free tieridentity emailcard?limit request-rateevidence documentedstale on 2026-12-01
- caveatPinecone Inference (embeddings and reranking)
Starter plan, labelled Free: 5,000,000 embedding tokens a month per model (llama-text-embed-v2, multilingual-e5-large, pinecone-sparse-english-v0...
free tieridentity emailcard?limit volume-quotaevidence documentedstale on 2026-12-04
- caveatPuter.js AI (user-pays)
Every account gets a free monthly allowance, shown in the dashboard; AI requests per 10 s are 20 for an anonymous (guest) account, 30 for a free account and...
free tieridentity anonymous?card?limit request-rateevidence documentedstale on 2026-11-28
- caveatQdrant Cloud Inference (free models on a free cluster)
BM25 and a dense model such as all-MiniLM-L6-v2 are free to use on free-tier clusters, with no token charge; larger models (mxbai, SPLADE) run on metered...
free tieridentity emailcard?evidence documentedstale on 2026-12-04
- caveatRequesty
Free plan: access to all free models at no token charge, 200 requests per day and 20 requests per minute for a new organization, counted across all free...
free tieridentity emailno cardlimit request-rateevidence documentedstale on 2026-11-28
- caveatRouteway
Models whose id ends in :free cost $0.00; limits are 5 requests per minute and 200 requests per day on the rate-limits page, and 20 requests per minute and...
free tieridentity cardlimit request-rateevidence documentedstale on 2026-12-05
- caveatToken Harbor
A $0 Free plan with a free allowance that resets on a rolling 7-day period, spent on :free model ids (DeepSeek V4.1 Flash, MiMo V2.6 Flash, TH-Rudder at the...
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-05
- caveatVercel AI Gateway
Every Vercel team account has a free tier with a monthly included credit usable on a subset of the catalogue, with a lower per-model rate limit than the...
free tieridentity cardlimit volume-quotaevidence documentedstale on 2026-12-05
- caveatWeaviate Embeddings (Weaviate Cloud free cluster)
Usable on a free Weaviate Cloud cluster at 2,000 embedding requests per cluster per day; with 200 objects per batch that is at most 400,000 embeddings a day.
free tieridentity emailcard?limit request-rateevidence documentedstale on 2026-12-04
- caveatZhipu BigModel Open Platform (bigmodel.cn)
Two models are free on input, output and cache, with no token cap or expiry stated: the text model GLM-4.7-Flash (200K context) and the vision model...
free tieridentity any-phoneno cardlimit request-rateevidence documentedstale on 2026-12-04
- caveatHetzner Inference API (experiment)
Free of charge while the Inference API remains in experimental status; two models, Qwen/Qwen3.6-35B-A3B-FP8 and Qwen3.8-27B, each with a 262,144-token...
free tieridentity cardphonelimit request-rateevidence reportedstale on 2026-11-19
- caveatNous Portal
Free plan at $0 per month with $0 in monthly credits: access to the models marked FREE in the catalog only, at standard rate limits; a third-party plan...
free tieridentity cardevidence reportedstale on 2026-11-19
Trial 26
- trialAI21 Studio (Jamba API)
New accounts are given a $10 credit good for three months on the AI21 platform, usable for the APIs, the SDK and the playground; valid billing information...
free tieridentity email?card?limit volume-quotaevidence documentedstale on 2026-11-28
- trialAkashML (Akash Chat API)
Trial credits are granted when a payment method is verified during onboarding; the amount is not stated on the current documentation pages.
identity cardlimit volume-quotaevidence documentedstale on 2026-12-05
- trialAssemblyAI (Inference)
USD 50 in free credits at signup, no credit card required; one-time, not a monthly allowance.
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-04
- trialBytePlus ModelArk free inference quota
A one-time free inference quota per model, 500K tokens per model, granted to new users on registration.
free tieridentity emailcard?limit volume-quotaevidence documentedstale on 2026-12-04
- trialCerebras
A Free Trial, not a free tier: USD 5 of credits granted after a verified payment method is added, expiring 30 days after they are granted.
identity cardlimit retention-clampedevidence documentedstale on 2026-11-28
- trialContextual AI reranker API
New signups receive USD 25 in free credits, USD 50 with a work email; the rerank API launch post states the first USD 50 as 1 billion tokens.
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-04
- trialCoze
A free quota of 30 Coze API calls for each registered account, counted over all APIs of all agents, then Coze tokens must be bought; separately 10 message...
free tieridentity email?card?limit volume-quotaevidence documentedstale on 2026-12-05
- trialDahl Inference
A one-time gift of 100 million tokens to the account pool at signup, with no monthly reset; a new key holds 0 tokens until tokens are allocated from the pool
free tieridentity anonymousno cardlimit volume-quotaevidence documentedstale on 2026-12-05
- trialDeepgram (Inference)
Free USD 200 credit on the Pay as you go plan with no minimums, no expiration and no credit card required; the vendor puts it at roughly 43,000 minutes...
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-04
- trialDeepL API (Developer plan)
A one-time credit of 1 million characters for text and document translation on the Developer plan: a total translation limit of 1,000,000 characters that...
free tieridentity emailcard?limit volume-quotaevidence documentedstale on 2026-12-04disagrees
- trialGrokified
A one-time $5 signup credit spent at 50% off list price, usable on chat, images and video; there is no recurring free tier
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-05
- trialInception Labs (Mercury API)
100 million free tokens assigned to each new account, no payment details required; free-tier limits are 1,000 API requests, 1,000,000 input tokens and...
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-11-28
- trialJina AI embeddings and reranker API
An auto-generated API key carries free tokens usable across every Jina model (embeddings, reranker, reader, classifier, segmenter).
free tieridentity anonymousno cardlimit volume-quotaevidence documentedstale on 2026-12-04
- trialMixedbread (managed search with reranking)
Starter plan: $5 in one-time credits, no card required; 3 workspace users, 3 stores, 100 queries per minute and 6,000 upload requests per minute.
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-04
- trialNebius Token Factory
No standing free allowance on the Starter Plan, which is pay-as-you-go.
free tieridentity cardlimit volume-quotaevidence documentedstale on 2026-12-04
- trialNVIDIA build
Free tier with no per-token billing, rate limited to up to 40 requests per minute (RPM) for most models; each account's own limit is shown in the dashboard
free tieridentity any-phoneno cardlimit request-rateevidence documentedstale on 2026-12-04
- trialReducto
USD 150 of free usage on the Standard plan, then list rates apply: Parse (r-1) USD 10 per 1,000 pages, Extract USD 20 per 1,000 pages
free tieridentity emailcard?limit volume-quotaevidence documentedstale on 2026-12-04
- trialScaleway Generative APIs (serverless)
Free Tier for Serverless: up to 1,000,000 tokens for models billed by tokens and 60 minutes of audio transcription for models billed by audio minutes...
free tieridentity cardlimit volume-quotaevidence documentedstale on 2026-12-05
- trialSpeechmatics (Inference)
USD 100 credit grant at signup on the Free plan, no card required; one credit is worth one US dollar, and one balance covers Speech to Text, add-ons and...
free tieridentity emailcardlimit volume-quotaevidence documentedstale on 2026-12-04
- trialStability AI (Inference)
25 free credits for a new account, where 1 credit equals USD 0.01, so the grant is USD 0.25 of API usage; one-time, and the getting-started guide ties the...
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-04
- trialStepFun Open Platform (platform.stepfun.ai)
No standing token allowance.
free tieridentity email?card?limit request-rateevidence documentedstale on 2026-12-04
- trialTencent HY Text Generation Global (Hunyuan-translation)
A one-time free resource package of 1M tokens for Hunyuan-translation only, shared consumption, valid for 1 year from service activation and deducted before...
free tieridentity national-idno cardphonelimit volume-quotaevidence documentedstale on 2026-12-04
- trialTetrate Agent Router Service
A one-time credit: USD 10 of free credits on the first automatic authentication with Tetrate through goose, for new and existing Tetrate users, per the...
free tieridentity cardlimit volume-quotaevidence documentedstale on 2026-12-04
- trialUnstructured Transform
10,000 free pages once per account, no card required, all features included; after the first 10,000 pages the API refuses further page processing unless the...
free tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-04
- trialVoyage AI (Inference)
The first 200 million tokens for voyage-4-large, voyage-4, voyage-4-lite, voyage-context-4 and voyage-code-4, or the first 50 million tokens for...
free tieridentity emailcardlimit volume-quotaevidence documentedstale on 2026-12-04
- trialXiaomi MiMo API Open Platform
One-time free credits for all new and existing users, collected on the console Balance page, announced with the start of billing.
free tieridentity emailno cardphonelimit volume-quotaevidence documentedstale on 2026-12-04
Paid only 20
20 hosted offers have no free allowance and are listed for their price, on Inference: paid only.