Portkey free tier includes Control panel for Gen AI apps featuring an observability suite & an AI gateway… Not verified — our last read, on 2026-10-24, found a change we could not reconcile. Compare with 33 alternatives in AI / ML.
Subtypes in AI / ML: model_gateway — one endpoint that reaches models run by several independent providers; what is sold is the routing, key management and fallback, not the weights; llm_observability — records the prompts, responses, latency, cost and tool calls an application sent to a model, for reading afterwards. We read that on 2026-09-01 from https://portkey.ai/, where it says: “Monitor LLM behavior, catch anomalies early” How we use this.
Portkey's free tier offers Control panel for Gen AI apps featuring an observability suite & an AI gateway… We are not rating this offer today; this page's Last read note says why.
Control panel for Gen AI apps featuring an observability suite & an AI gateway. Send & log up to 10,000 requests for free every month.
No recorded pricing changes for Portkey. We publish no change we refused to record, so treat the empty history as a statement about our records, not about this vendor's pricing.
How Portkey compares against every one of the 33 we list, not against a selection of them. A surface with room for every member of a tied set may rotate their order. A surface with room for only some of them names none of them, because which ones it named would be the claim, and a claim seeded on the date is not the one we serve tomorrow. How we decide this.
| Service | Free Tier | Stability |
|---|---|---|
| Portkey | Free | unrated |
| Pollinations.AI | Free Keyless Classic API: one image model and GPT-OSS 20B text, free, no sign-up, rate-limited per IP. Every other model needs an API key and Pollen, and free Pollen comes only from quests. |
caution |
| NVIDIA NIM | Free Free inference endpoints on build.nvidia.com: models marked Free Endpoint, including DeepSeek V4.1 Flash, GLM 5.3 Flash, gpt-oss-20b, Gemma 4 31B and NVIDIA Nemotron 3 Ultra, can be called at no cost. Their model IDs are deepseek-ai/deepseek-v4.1-flash, z-ai/glm-5.3-flash, openai/gpt-oss-20b,… |
stable |
| Maxim AI | Free Simulate, evaluate, and observe your AI agents — The 'OSS' tier is free forever for developers, small teams, and self-managed deployments. It includes features like a drop-in replacement, OTel compatibility, metrics & traces, built-in observability, budget management, and more. |
stable |
| Google Gemini API | Free (Reduced) Free tier: Gemini 4 Hydrogen, Gemini 3.8 Flash, 3.6 Flash, 3.5 Flash-Lite, 3.1 Flash-Lite, Gemini 3 Flash Preview, Gemini Embedding 2 and Gemma 4 are free of charge. Also free of charge are Gemini 3.8 Live (real-time voice), Gemini 3.5 Live Translate (translation), Gemini 3.5 Transcribe… |
stable |
| Cloudflare Workers AI | Free Workers AI runs AI models on Cloudflare's global network. Every account gets 10,000 Neurons per day at no charge, reset daily at 00:00 UTC; on the Workers Free plan, requests beyond that fail until the reset. |
caution |
| LangSmith | Free LLM observability and evaluation platform by LangChain — Developer plan is $0/seat/month with up to 5k base traces/month, then pay-as-you-go. LCU is $1.50/unit and LSU is $1.00/unit. |
unrated — change unconfirmed |
| Braintrust | Free Evals, prompt playground, and data management for Gen AI. Starter plan (free): 1 GB processed data/month, 10k scores/month, 14 days data retention. |
stable |
| Lumenfall.ai | Free AI media gateway: one OpenAI-compatible API to image and video generation models from several providers, billed at each provider's price with no markup, platform fee or subscription. One model, FLUX.1 [schnell] FP8, is listed free per image ("Free to try", served through Fireworks AI). |
stable |
| LLM7.io | Free LLM inference gateway. With a free token from dash.llm7.io: 1 request a second, 60 a minute, 250 an hour and 100,000 tokens per 24 hours, counting input and output tokens; LLM7 says free quotas may be reduced without notice. |
caution |
| Hugging Face | Free ML model hub. Since 2026-10-07, free accounts get no included Inference Providers credits: free users buy credits or upgrade to PRO to call Inference Providers, which serves 200+ models. |
caution |
| Groq | Free Fast LLM inference on Groq's LPU hardware. Free plan: gpt-oss-20b, gpt-oss-120b and Qwen3.8 27B, each at 30 requests a minute, 1,000 requests and 200,000 tokens a day, plus Whisper speech-to-text at 2,000 requests a day. |
caution |
| Keywords AI | Free Rebranded to Respan; keywordsai.co redirects to respan.ai. Free plan: full platform, 100k logs, 1k scores, 5 datasets, 2 evaluators, 5 prompts. |
source unconfirmed |
| Arize AX | Free Platform for agent observability, evaluation and improvement. AX Free costs $0 with no credit card required: 25k trace spans per month, 1 GB ingestion per month, 15 days of retention, unlimited users and evals, 10 Signal issues per month, the Alyx agent and community support. |
unrated |
| OpenRouter | Free AI model router. Free plan: 50 requests a day, no BYOK. |
stable |
| Langfuse | Free Open-source LLM engineering platform for tracing, evaluating and debugging AI applications. The Langfuse Cloud Hobby plan is free with no credit card: 50k units a month (each trace, observation and score counts as one unit), 30 days of data access, 2 users, community support via GitHub, and all… |
stable |
| Cohere | Free AI model API for Command chat models, Embed, Rerank, Transcribe and Parse. Every account starts with a Trial API key: calls made with it are free and limited to 1,000 API calls a month. |
stable |
| Hetzner Inference API | Free Hetzner's experimental Inference API: an OpenAI-compatible API for open-weight models, free while it stays experimental. Base URL: https://inference.hetzner.com/api/v1. |
stable |
| Comet ML | Free Comet makes Opik, for LLM and agent observability and evaluation, and an MLOps platform for ML experiment tracking. Opik is open source and free to self-host. |
stable |
| SiliconFlow | Free Pay-as-you-go model inference. siliconflow.cn, run by a Beijing company, lists these models at no charge: the chat models Qwen3-8B, Qwen2.5-7B-Instruct (Free), Qwen3.5-4B, GLM-4-9B-0414, GLM-Z1-9B-0414, DeepSeek-R1-0528-Qwen3-8B (Free) and Xing4.0-29B; the translation model Hunyuan-MT-7B; the OCR… |
unrated |
| LangWatch | Free A LLMOps platform helping AI teams measure, monitor, and optimize LLM applications for reliability, cost-efficiency, and performance. With a powerful DSPy component, we enable seamless collaboration between engineers and non-technical teams to fine-tune and productionize GenAI products. |
unrated |
| Zhipu AI | Free China-based GLM model API. The price list marks these models free: GLM-4.7-Flash, GLM-4-Flash-250414 and GLM-Z1-Flash (text); GLM-4.6V-Flash, GLM-4V-Flash and GLM-4.1V-Thinking-Flash (vision); CogView-3-Flash (image); CogVideoX-Flash (video). |
source unconfirmed |
| Fireworks AI | Free Credits Inference platform for open models, with serverless per-token pricing and on-demand GPU deployments billed per GPU second. Fireworks offers $1 in free credits to get started. |
unrated |
| Baseten | Basic (Free Credits) ML model deployment platform. New workspaces receive credits for testing and deployment; Baseten does not state the amount. |
unrated |
| Cerebras | Trial Ultra-fast LLM inference API — no permanent free tier. New accounts get $5 in free credits after adding a verified payment method, expiring 30 days after they are granted. |
risky |
| GitHub Models | Retired GitHub retired GitHub Models on 2026-07-30, so there is no free tier. GitHub's own documentation states "GitHub Models has been retired." |
source unconfirmed |
| Anthropic API | Pay-as-you-go Claude API with usage-based pricing per million tokens (input/output). Claude Fable 5.2 $8/$40. |
unrated |
| Mistral AI | Pay-as-you-go Since about 2026-11-05, Mistral's Free plan covers Vibe and Studio and does not include API access. Creating API keys and calling the API require billing activation and pay-as-you-go with prepaid credits. |
unrated |
| Together AI | Pay-as-you-go Inference API for open-source models, billed per token. Together AI does not currently offer free trials; access requires a minimum $15 credit purchase for an organization's first purchase ($5 for later purchases), and the platform is fully prepaid. |
unrated |
| MiniMax | Pay-as-you-go MiniMax sells API access pay-as-you-go on platform.minimax.io. Its pricing pages list no free API allowance; one docs page mentions a free quota for new pay-as-you-go users but gives no amount. |
unrated |
| Ollama Cloud | Pay-as-you-go Cloud-hosted Ollama for running open-source LLMs. Between October 26 and 27, 2026, Ollama replaced its Free plan, which included starter usage credits, with a $0 Pay as you go plan. |
unrated |
| DeepSeek API | Pay-as-you-go Pay-as-you-go LLM API with no free tier, billed from a prepaid balance. DeepSeek-V4.1-Flash (deepseek-flash): $0.30/M input, $1.20/M output at peak. |
unrated |
| xAI | Pay-as-you-go Grok API, pay as you go: sign up at console.x.ai, then load it with credits. grok-4.7 $2.00/$6.00 (per 1M tokens, under 200k prompt tokens). |
unrated |
| OpenAI | Pay-as-you-go AI API platform. One model is priced Free in OpenAI's own table: the moderation model omni-moderation-latest. |
unrated |
Left out of this list: Brave Search API (shares no subtype with this product), Pinecone (shares no subtype with this product), Qdrant (shares no subtype with this product), Deepgram (shares no subtype with this product), Kaggle (shares no subtype with this product), Roboflow (shares no subtype with this product), Scale AI (shares no subtype with this product), AssemblyAI (shares no subtype with this product), Replicate (shares no subtype with this product), Weights & Biases (shares no subtype with this product), Clarifai (we have not classified this record), Neptune.ai (we have not classified this record), Labelbox (shares no subtype with this product), Google Colab (shares no subtype with this product), Composio (shares no subtype with this product), DeepAR (none of this category's subtypes apply), OCR.Space (shares no subtype with this product), Parseur (shares no subtype with this product), Reducto (shares no subtype with this product), SerpApi (shares no subtype with this product), Tavily AI (shares no subtype with this product), paperspace (shares no subtype with this product), Audio Enhancer (none of this category's subtypes apply), Clair (none of this category's subtypes apply), Langtrace (we have not classified this record), Mediaworkbench.ai (none of this category's subtypes apply), Othor AI (none of this category's subtypes apply), ReportGPT (none of this category's subtypes apply), Zenable (none of this category's subtypes apply), Exa (shares no subtype with this product), GPU-Bridge (we have not classified this record), Google Vertex AI Agent Engine (we have not classified this record), Google Vector Search 2.0 (shares no subtype with this product), Google Gemini Embedding 2 (shares no subtype with this product), Alibaba Cloud Qwen Code (none of this category's subtypes apply), E2B (shares no subtype with this product), Modal (shares no subtype with this product), ElevenLabs (we have not classified this record), Stability AI (we have not classified this record), Daytona (shares no subtype with this product), Blaxel (shares no subtype with this product), Novita Sandbox (shares no subtype with this product), Runloop (shares no subtype with this product), Mem0 (shares no subtype with this product), Zep (shares no subtype with this product), Letta (we have not classified this record), Supermemory (shares no subtype with this product), Cognee (shares no subtype with this product), Serper (shares no subtype with this product), Linkup (shares no subtype with this product), Parallel (shares no subtype with this product), Jina AI (shares no subtype with this product), You.com (shares no subtype with this product), Cloudflare AI Search (we have not classified this record), Valyu (shares no subtype with this product). Each is still listed in AI / ML, with the vendor's own words we read that from on its page. How we decide this.
We do not push notifications. Every change we record is readable the moment it lands, and the JSON answer carries an ETag — send it back as If-None-Match and an unchanged answer costs one 304.
curl https://agentdeals.dev/api/changes?vendor=Portkey
curl https://agentdeals.dev/pricing-changes/feed.xml?vendor=Portkey