FAQ
Evaluating the product
What is TinyFish and what problem does it solve?
TinyFish is enterprise infrastructure for AI web agents that gives AI models live, authenticated access to the web. It replaces cached, stale results with real-time data retrieval, browser automation, and multi-step web workflow execution at scale. It handles everything from proxy rotation and anti-bot handling to credential management, so developers don't need to manage underlying browser clusters.
What makes TinyFish different from other web scraping or automation tools?
TinyFish offers a single, integrated platform with a unified API and credit pool that replaces fragmented scraping and automation tools. Unlike CSS-selector-based tools that break when sites change, TinyFish uses semantic element finding so workflows remain stable. It also natively handles authenticated flows, dynamic pages, and vendor or government portals that have no standard API.
Who uses TinyFish?
TinyFish serves enterprise customers including Google Hotels, DoorDash, ClassPass, Grubhub, Digital Garage, The Zebra, and TestSprite, as well as unnamed Fortune 500 companies across hospitality, insurance, retail, and logistics. The platform runs hundreds of thousands of enterprise web agents every month, executing over 35 million operations monthly as of August 2025.
What are real customers saying about TinyFish?
Jonathan Lam, Senior Program Manager at ClassPass, says TinyFish 'solves real problems and makes a measurable impact.' Abhi Shah, Director of Data Science at DoorDash, notes that TinyFish's platform 'manages web interaction complexity at scale.' Yuto Saeki, Head of AI Strategy at Digital Garage, highlights that TinyFish enabled web workflows in Japan that couldn't be automated any other way.
Is TinyFish suitable for large-scale enterprise workloads?
Yes. TinyFish is purpose-built for enterprise scale, executing 35M+ operations per month as of its August 2025 Series A announcement. The platform supports hundreds of concurrent browser sessions and agent runs on Enterprise plans, with custom rate limits and concurrency available. It is backed by $47M in Series A funding led by ICONIQ.
Understanding pricing and plans
How does TinyFish pricing work?
TinyFish uses a credit-based model: 1 credit equals 1 agent step or 4 minutes of browser time. LLM inference, stealth browser, anti-bot, and proxy are all included in that single credit at no additional charge. Search and Fetch APIs are free on every plan with no credit consumption.
What plans are available?
TinyFish offers four tiers: Pay As You Go ($0/month, $0.015/credit), Starter ($15/month, 1,650 credits included, $0.014/credit overage), Pro ($150/month, 16,500 credits included, $0.012/credit overage), and Enterprise (custom pricing and rate limits). Search and Fetch APIs are free on all tiers.
How does TinyFish pricing compare to alternatives?
At 1,000 tasks per month, a competing stack costs approximately $2,400 versus $15 on TinyFish — a roughly 160x cost difference. At 10,000 tasks per month, the comparison is $24,000 versus $150. TinyFish consolidates what would otherwise require multiple tools into a single platform with a single credit pool.
What happens if I run out of credits?
On Starter and Pro plans, workflows continue running at the overage rate so there are no mid-run failures. You can set a spending cap in the dashboard if you want to limit overage. On Pay As You Go, you pay per credit as you go, so there is nothing to run out of.
Are failed runs billed?
No. Failed runs cost $0 on every plan, including Pay As You Go, Starter, Pro, and Enterprise. This is listed as a platform-wide included feature.
Understanding the product and APIs
What APIs does TinyFish offer?
TinyFish provides four core APIs: Search API (fast structured web search, free on all plans), Fetch API (converts any URL to clean content, free on all plans), Agent API (multi-step web automation, billed at 1 credit per step), and Browser API (cloud browser sessions, billed at 1 credit per 4 minutes, capped at 60 minutes per session). All plans also include Mako, TinyFish's web-native model built for live web workflows.
What is the Mako model?
Mako is TinyFish's web-native model built specifically for live web workflows. It is included as part of the TinyFish platform and is designed to navigate, authenticate, and execute complex multi-step tasks on the live web rather than relying on cached data.
What infrastructure complexity does TinyFish manage for developers?
TinyFish manages browser clusters, proxy rotation, anti-bot handling, credential management, LLM inference, and structured output generation — all abstracted away from the developer. Developers interact through a single API key and credit pool without needing to provision or maintain any browser infrastructure.
Does TinyFish work with authenticated and dynamic web content?
Yes. TinyFish is explicitly designed to handle authenticated web workflows, dynamic pages, and portals — including vendor and government sites — that have no standard API. This is a core differentiator versus tools limited to public, static web content.
What is the Browser API session time limit?
Browser API sessions are billed at 1 credit per 4 minutes of browsing, with a 60-minute cap per session. Concurrent session limits depend on plan: 5 on Pay As You Go, 10 on Starter, 50 on Pro, and custom on Enterprise.
Getting started and integrating
How do I get started with TinyFish?
You can sign up for free at agent.tinyfish.ai. The Search and Fetch APIs are free on every plan with no credit requirement, making them a zero-cost entry point. Documentation is available at docs.tinyfish.ai, and open-source examples are available at github.com/tinyfish-io/tinyfish-cookbook.
Does TinyFish provide documentation and code examples?
Yes. Full API documentation is available at docs.tinyfish.ai. TinyFish also maintains an open-source cookbook of examples at github.com/tinyfish-io/tinyfish-cookbook to help developers get started quickly with real-world automation patterns.
Which plan should I start with if I'm just evaluating?
Pay As You Go is ideal if usage is unpredictable or you're just starting — there's no monthly commitment and you pay $0.015 per credit only when you use it. The Search and Fetch APIs are free on this tier, so initial exploration costs nothing. Upgrade to Starter or Pro once you're running workflows daily and want the 40% cost reduction.
Is there an accelerator program for teams building AI web agents?
Yes. The TinyFish Accelerator is a 9-week, fully virtual program with rolling admissions, backed by a $2M seed pool from Mango Capital and 15+ partners. It covers applying, building with partner stack access and weekly mentorship, and pitching on Demo Day for funding decisions. Contact Robin Vasan, Managing Partner at Mango Capital, for accelerator inquiries.
Assessing reliability and scale
How reliable is TinyFish at scale?
TinyFish executes over 35 million operations per month as of August 2025 for enterprise customers across multiple industries globally. Yunhao Jiao, CEO of TestSprite, notes that 'TinyFish handles browser interaction with the reliability our testing cycles demand.' The platform is backed by $47M in Series A funding led by ICONIQ to support continued infrastructure scaling.
How does TinyFish handle speed at scale?
The Chief AI Officer of The Zebra reports that 'workflows that once took minutes now happen in seconds' with TinyFish, and that 'speed compounds at scale.' The platform is designed to run hundreds of browsers simultaneously using a serverless architecture, eliminating the bottleneck of managing browser clusters manually.
Does TinyFish work for international or region-specific web workflows?
Yes. TinyFish has documented enterprise deployments in Japan and across the US. Digital Garage's Head of AI Strategy specifically noted that TinyFish enabled web workflows in Japan that couldn't be automated any other way, including complex authenticated and regional portal access.
What funding and backing does TinyFish have?
TinyFish raised $47 million in Series A funding in August 2025, led by ICONIQ with participation from USVP, Mango Capital, MongoDB Ventures, ASG, and Sandberg Bernthal Venture Partners. The company is led by Co-founder and CEO Sudheesh Nair, who stated: 'The browser is where the future battles of AI will be fought. We're building the infrastructure for that fight.'