Back to vendors
T

TinyFish

Also known as: Tinyfish

Visit site
Entry pricePay per use: Agent $0.016 a step, Browser $0.002 a minute, Monitor $0.005 a check; Search and Fetch freeFull pricing detail

Enterprise infrastructure for AI web agents that navigate, authenticate, extract, and transact across hundreds of live sites in parallel through a single Model Context Protocol native API.

TinyFish is enterprise infrastructure for AI web agents, founded in 2024 in Palo Alto and led by chief executive Sudheesh Nair, and it launched with forty seven million dollars in funding led by ICONIQ Capital. The premise is that much of the web's most valuable data sits behind logins, forms, and paywalls where search engines cannot reach and traditional automation fails. TinyFish deploys swarms of specialized agents that sign in, navigate, extract, and act like humans at scale, and already runs hundreds of thousands of enterprise web agents every month for customers including Google, DoorDash, Amazon, and ClassPass across hospitality, insurance, retail, and logistics.

The platform brings five capabilities under one application programming interface key and one prepaid wallet. Search returns structured results from live pages rendered by real browsers rather than cached indexes, Fetch converts any URL into clean markdown or JSON, Browser provides stealth sessions that defeat anti bot protection, and Agent runs autonomous multi step web workflows. Monitor watches a page or a search topic on a schedule and sends an email or webhook when it changes. It is serverless, with no browsers to manage or proxies to configure, and can run up to a thousand simultaneous operations across hundreds of sites, completing in minutes what would take days by hand.

TinyFish is Model Context Protocol native, so it works with Claude, Cursor, and any compatible client, and it also offers a direct application programming interface, a command line tool and integrations with tools like n8n, Dify and Vercel. Its open source AgentQL project adds an AI powered query language with Python and JavaScript software development kits, a REST endpoint, and a browser based debugger, using natural language selectors that self heal as sites change. Agents keep logins and site state across sessions through saved profiles and vaults, and return structured outputs to downstream systems and webhooks, with a screenshot and an HTML snapshot of every step.

The platform is built for enterprise governance, emphasizing consistency, inspectability, and control through deterministic web execution and structured, auditable artifacts. Pricing is pay as you go from a prepaid wallet, at $0.016 an agent step, $0.002 a browser minute and $0.005 a monitor check, with Search and Fetch free and $8 of credit on a new account. Its Enterprise plan adds ISO 27001, enterprise SSO, audit logs, VPC deployment and a 99.99 percent uptime commitment beside the serverless cloud service. TinyFish names no persistent learning memory and offers no user facing model selection, since it deliberately bundles model inference into one usage based price.

From our comparison video

TinyFish vs Browserbase (2026): Which Should You Choose?

Our 3 minute video puts TinyFish side by side with Browserbase. It covers what each one does and how each is priced, then when to choose which.

Recorded in October 2026. The profile here is live and updates as research is verified.

Vendor details

Canonical URL

https://tinyfish.ai

Category

Browser / computer-use agent

Company status

independent

Use cases & customers

In practice

An energy company must pull invoices from hundreds of vendor portals. TinyFish agents log into 341 portals across 1,643 monthly logins, navigate multi step workflows, and return structured invoices and status updates to downstream systems automatically.

A small hotel in Japan with eight rooms and no API needs its availability on Google. A TinyFish agent signs into the booking system, reads dates and pricing, and updates the Google Hotels listing without the hotel changing anything.

An insurance marketplace collects quotes from many carriers. TinyFish handles each provider's different authentication and form structure, retrieves rates, and returns structured, comparable quotes, running many portals in parallel in a fraction of the manual time.

Agentic Index coverage score

7.5 / 14 capabilities · 54%

Integrations & Tool Calling Partial

Agents act on websites through the browser and log in with credentials from a connected 1Password or Bitwarden vault, and results go to webhooks. The MCP, n8n and Dify integrations let other tools call TinyFish. TinyFish describes no connector catalog or custom tool support for the agent. Credentials never enter a prompt, and agents handle SSO and MFA logins without reading the password.

SourceTinyfish, docs.tinyfish.ai Connect Your Vault and llms.txt index, and tinyfish.ai/enterpriseread 2026-10-05

Workflow Orchestration Partial

Each run is one goal the agent carries out over many steps, and many runs start in parallel through the bulk asynchronous API. Sequencing and branching across runs are left to the customer's code, and TinyFish describes no workflow model. Runs can be blocking, asynchronous or streamed as server sent events, and can be cancelled midway.

SourceTinyfish, docs.tinyfish.ai Agent API referenceread 2026-10-05

Knowledge Grounding & RAG Not documented

Search and Fetch read the live web at run time, and TinyFish describes no maintained store of the customer's own knowledge. The Research API searches, weighs sources and writes a cited report from a question in plain language.

SourceTinyfish, docs.tinyfish.ai Research APIread 2026-10-05

Human Oversight & Guardrails Partial

Customers can set constraints on a run. A strict mode stops instead of exploring workarounds, and runs take a step limit of 1 to 500 and a maximum duration (step limit and strict mode are in beta). TinyFish describes no approval step or pause for a person. Runs can be cancelled while in progress.

SourceTinyfish, docs.tinyfish.ai Agent API referenceread 2026-10-05

Security, Identity & Governance Full

The Enterprise plan carries ISO 27001, enterprise SSO and audit logs, with role based access, encrypted credential handling and a Vanta trust center at trust.tinyfish.ai. Credentials stay in the customer's own vault, can be revoked at any time and never enter a prompt.

SourceTinyfish, tinyfish.ai/pricing and /enterpriseread 2026-10-05

Observability & Auditability Full

Runs can be listed and searched. Each step keeps a screenshot and an HTML snapshot through the API, runs stream progress and a live browser preview, and the Enterprise plan adds audit logs. Every action produces a screenshot and a replay link for audit, Pay as you go includes storage for screenshots and logs, and Monitor keeps a run history.

SourceTinyfish, tinyfish.ai/pricing, /enterprise and docs.tinyfish.ai Monitorread 2026-10-05

Memory & State Persistence Not documented

Browser context profiles keep logins and site state across runs, and Monitor keeps a history of runs. TinyFish describes no memory the agent reads to decide. Disconnecting a vault removes its cached credentials at once.

SourceTinyfish, docs.tinyfish.ai llms.txt index, Monitor and Connect Your Vaultread 2026-10-05

Deployment & Data Residency Full

The Enterprise plan offers VPC deployment beside the serverless cloud service, and data residency and access scope are agreed in the contract. No region list is published. Enterprise also carries a 99.99% uptime SLA.

SourceTinyfish, tinyfish.ai/pricing and /enterpriseread 2026-10-05

Prebuilt Agents, Templates & Packs Not documented

Customers get code examples (scraping, form filling, bulk requests) and a prompting guide, but no packaged agents or templates to adopt. TinyFish's products are separate APIs (Agent, Search, Fetch, Browser and Research), not ready made agents, and a public cookbook on GitHub collects more examples.

SourceTinyfish, docs.tinyfish.ai llms.txt indexread 2026-10-05

Triggers & Channel Coverage Full

Monitor runs checks of a page or web topic on a recurring schedule and sends changes to a webhook or email, beside runs started from the API, CLI, MCP, n8n and Dify. Topic monitors run at most every 30 minutes, and each completed monitor run costs $0.005.

SourceTinyfish, docs.tinyfish.ai Monitorread 2026-10-05

Model Flexibility & Routing Not documented

The Agent API has no model setting. LLM inference is bundled into the per step price of $0.016 with no key to bring, and TinyFish's Mako is its own model.

SourceTinyfish, docs.tinyfish.ai Agent API reference and tinyfish.ai/pricingread 2026-10-05

APIs, SDKs & MCP Extensibility Full

A REST API covers agent runs, runs, browser sessions, search, fetch, monitors, vaults and profiles, alongside a CLI and an MCP server. AgentQL is a separate product from the same company. The MCP server at agent.tinyfish.ai/mcp signs in with OAuth 2.1 and exposes automation, fetch, search, browser session, batch and monitor tools to clients such as Claude, ChatGPT, Cursor and Codex, and a Research API returns cited reports.

SourceTinyfish, docs.tinyfish.ai MCP Integration and llms.txt indexread 2026-10-05

Testing, Debugging & Optimization Not documented

The benchmarks are TinyFish's own comparison, and step snapshots record each run. TinyFish describes no evaluation harness, scored test cases or quality gate for the customer's agent runs. Each run reports how many steps it took and ends as completed, failed or cancelled.

SourceTinyfish, docs.tinyfish.ai llms.txt index and Agent API referenceread 2026-10-05

Browser & Computer Use Full

The Agent API takes a goal and a URL and navigates, authenticates, fills forms and extracts across live sites in lite or stealth browsers, and the Browser API provides remote browser sessions with a live preview. Browsers run with residential proxies and anti bot handling, and browser sessions cost $0.002 a minute.

SourceTinyfish, docs.tinyfish.ai Agent API reference and tinyfish.ai/pricingread 2026-10-05

The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded

Product updates

·Agent capabilityVerified

Tinyfish launched Mako, a web-agent-native AI model built to operate the live web and execute multi-step workflows at scale. Trained on real production data from authenticated enterprise tasks, the model is now the default engine behind the Tinyfish Web Agent. Mako natively encodes web elements and caches page state to maintain context across long workflows without exceeding context limits.

Bears on: Agent capability

View source
View all TinyFish updates →Tracked since Jul 2026 · Verified from public vendor sources

Pricing

Pay per use: Agent $0.016 a step, Browser $0.002 a minute, Monitor $0.005 a check; Search and Fetch free

Free tierTrial available

What is public

TinyFish prices by use, with no plans and no monthly minimum. Agent costs $0.016 per step, Browser $0.002 per minute and Monitor $0.005 per check, drawn from prepaid wallet funds added in deposits of $10 or more, and a new account starts with $8 of credit. Search and Fetch are free within rate limits at any balance. Vaults, profiles, the SDKs, the command line tool and the MCP server come with pay as you go. Enterprise is a contract that adds ISO 27001, enterprise SSO, audit logs, VPC deployment, a 99.99 percent uptime commitment and limits set to the workload. The earlier Starter $15 and Pro $150 plans and the $0.015 per step rate are withdrawn.

Free / trial

$8 of wallet credit on a new account, and Search and Fetch free within rate limits

Lowest paid plan

No plans; prepaid Wallet

Key ambiguities

Steps per task vary with the site and goal, so cost per task is estimated rather than fixed.

Agentic Index verified 2026-10-21

Alternatives to TinyFish

The closest documented capability profiles to TinyFish among browser and computer-use agents tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.

  • Notte9.5 / 14Adds documented Model Flexibility & Routing and Testing, Debugging & Optimization
  • Autotab8.0 / 14Adds documented Memory & State Persistence and Testing, Debugging & Optimization
  • H Platform9.0 / 14Adds documented Prebuilt Agents, Templates & Packs and Testing, Debugging & Optimization
  • AGI, Inc.6.0 / 14Adds documented Memory & State Persistence
  • Browserbase10.0 / 14Adds documented Memory & State Persistence and Model Flexibility & Routing, among othersTinyFish vs Browserbase →Video
  • AgentQL5.5 / 14Adds documented Model Flexibility & Routing and Testing, Debugging & Optimization

Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded

Head to head

Contact us

Found a vendor we missed? Have feedback on the index? We'd love to hear from you.