How much does Bika.ai cost?

Free covers up to 5 seats with 3,000 daily AI credits, 5 GB of storage, 10,000 records per database and 200 automation runs. Plus is $19 per seat a month or $16 billed yearly with 50,000 AI credits per seat and 30,000 automation runs; Team is $39 or $33 with 100,000 credits per seat and 60,000 runs; Business is $199 or $166 with 500,000 credits per seat, 100,000 runs and audit logs. Paid plans currently add an equal number of bonus credits as a limited offer. Subscription credits refresh each billing date and do not roll over. Self hosted and enterprise licenses are sold through sales.

Does Bika.ai support MCP and tool calling?

Full under the Agentic Index verification standard. Bika.ai publishes public evidence of Model Context Protocol (MCP) support, tool calling, and API or SDK extensibility for connecting agents to external tools. Evidence basis: A REST OpenAPI with a published OpenAPI specification and an object oriented TypeScript SDK reaches spaces, nodes, databases, records and automations with bearer token authentication, with cURL, Python and TypeScript examples, alongside an. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Can Bika.ai be self hosted or run in a private VPC?

Full under the Agentic Index verification standard. Bika.ai publishes public evidence of self hosted or private VPC deployment, along with data residency options for controlling where agent data is processed and stored. Evidence basis: Besides the hosted service, Bika sells a self hosted edition installed on the customer's own servers for data sovereignty and compliance, requested through a license form that returns a deployment guide, an installation package and a 30 day. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Can you bring your own model or route between providers with Bika.ai?

Full under the Agentic Index verification standard. Bika.ai publishes public evidence of bring your own model and bring your own key (BYOK), and of routing tasks between models or providers rather than being locked to one. Evidence basis: Each agent picks its model, and the default Auto setting lets the system choose the best one and moves to newer models as they arrive. Models cost credits by a published multiplier, so a heavier model costs about twice a baseline one and a. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Back to vendors

Bika.ai

Q: Who is Bika.ai for?

one person companies and solo founders, content creators and social media managers, small marketing, sales and operations teams, product and project teams tracking tickets and requirements, teams moving off Airtable or AITable that want agents alongside their data

Also known as: Bika, bika.ai, Bika AI

Visit site

Agent builderindependentVerified 2026-10-18

Entry pricePlus $19 per seat a month, or $16 a month billed yearlyFull pricing detail

Bika.ai is a workspace for running a small company with a team of AI agents, combining agents with skillsets and MCP servers, automations, spreadsheet style databases, dashboards and documents in one space. Agents read and write the space's data, publish posts and call outside APIs, automations run on schedules, webhooks, emails and record changes, and 145 installable templates package ready made agentic apps.

Bika.ai sells what it calls an AI organizer for one person companies and small teams: a single space where AI agents, automations, databases with billions of rows, forms, dashboards, documents, missions and reports work together, with agents listed in the team directory like digital employees. Templates cover marketing, sales, finance, project management and operations, from a Twitter assistant and an email marketer to a lightweight CRM and invoice reminders.

An agent has a model, a prompt, a set of data sources it may use (databases, documents, automations and folders) and skillsets that let it read and write databases, create documents, generate images, call third party APIs and connect to more than 2,000 MCP servers through ToolSDK.ai or to a custom MCP server over Streamable HTTP.

Agents run tasks in the cloud while users switch between several at once, keep conversation history that can be shared by link, and can be called from an automation.

Automations start on a schedule, an incoming webhook, a new or matching record, a date field, a form submission, an inbound email or a change in an HTTP response, and chain conditions, loops, delays, filters, round robin assignment, scripts, HTTP requests, emails and messages to Slack, Telegram, WeCom, Feishu and DingTalk.

The Free plan covers 5 seats with 3,000 daily AI credits and 200 automation runs. Plus costs $19 per seat a month or $16 billed yearly, Team $39 or $33, and Business $199 or $166, with AI credits, storage, records and automation runs rising by tier; Team adds organization permissions and advanced API access, and Business adds audit logs and advanced permissions. Credits cover AI inference, with models priced by a multiplier. Access control works through roles, teams, guests on a whitelist and per resource permissions, there is no single sign on or published security certification, and a self hosted licensed edition with a 30 day trial and white label options is available through sales.

Bika.ai fits solo founders, creators and small teams that want agents, data and automations in one place, with templates to start from and the option to self host. It fits less well for larger companies that need single sign on, certifications, detailed run tracing or evaluation, or agents that operate a browser.

Vendor details

Canonical URL

https://bika.ai

Subcategory

All in one workspace for small teams combining AI agents with skillsets and MCP servers, automations, databases, dashboards and installable agentic app templates

Funding status

Privately held company.

Company status

independent

Use cases & customers

Primary use cases

drafting and publishing social posts and scheduling YouTube uploadsoutbound email sequences and follow upslightweight CRM, order and contractor management with automationsscheduled reports and notifications to WeCom, Slack or Telegramissue and ticket tracking with AI triageinvoice and expense collection with reminders

Target customers

one person companies and solo founderscontent creators and social media managerssmall marketing, sales and operations teamsproduct and project teams tracking tickets and requirementsteams moving off Airtable or AITable that want agents alongside their data

Deployment options

SaaSSelf hosted licensed edition with a 30 day trial, white label available

Integrations

Agents use skillsets for databases, documents, images and third party APIs, more than 2,000 MCP servers through ToolSDK.ai, and custom MCP servers over Streamable HTTP. Automations connect to GitHub, X (Twitter), Zoom, Slack, Telegram, WeCom, Feishu, DingTalk, IMAP and SMTP email, MySQL, PostgreSQL, Airtable, AITable, APITable, AWS Textract, Make and Zapier, plus webhooks, HTTP requests and scripts. Models include OpenAI, Amazon Bedrock, Azure AI, Anthropic Claude, Google AI, DeepSeek, Alibaba Qwen and ByteDance Doubao. A REST OpenAPI with a TypeScript SDK covers spaces, nodes, databases and records.

In practice

A solo founder installs the Email Marketer template, imports leads into its database, and lets an automation send a first email and two follow ups three days apart, while an agent drafts each message from the lead's record.

A creator chats with a Twitter agent, shares a screenshot of an article, and the agent searches the web, drafts a tweet and publishes it through the X integration once she approves the draft.

An operations lead sets a scheduler automation that every weekday pulls overdue tasks from a database, has an agent summarize them, and posts the summary to the team's WeCom group through its webhook.

Sources & related URLs

Related / legacy domains

https://bika.ai/help https://bika.ai/pricing https://bika.ai/template https://bika.ai/help/guide/developer/openapi

Research sources

https://bika.ai/ https://bika.ai/llms.txt https://bika.ai/pricing https://bika.ai/help/guide/ai/ai-agent https://bika.ai/help/reference/node-resource/ai https://bika.ai/help/reference/automation-action/toolsdk-ai https://bika.ai/help/reference/automation-action/call-agent https://bika.ai/help/reference/automation-trigger/features-list https://bika.ai/help/reference/node-resource/automation https://bika.ai/help/reference/integration/features-list https://bika.ai/help/guide/space/resource-permission https://bika.ai/help/guide/space/team-management https://bika.ai/help/guide/space/credit https://bika.ai/help/guide/space/intelligence https://bika.ai/help/guide/self-hosted/quick-start https://bika.ai/help/guide/developer/openapi https://bika.ai/sitemap/en/templates-sitemap.xml

Agentic Index coverage score

10.0 / 14 capabilities · 71%

Integrations & Tool Calling	Full
Agents act through skillsets that read and write the space's databases, create documents, generate images and call third party APIs, and through more than 2,000 MCP servers offered via ToolSDK.ai, plus any custom MCP server with a public Streamable HTTP endpoint. Ready skillsets cover Google Analytics, stock market data and X (Twitter), where an agent can draft and publish a post. Automations add actions for GitHub, X, Zoom, IMAP and SMTP email, MySQL and PostgreSQL, Slack, Telegram, WeCom, Feishu and DingTalk messages, HTTP requests and JavaScript or Python scripts, so agent output can be written to records and pushed to outside systems. Sourcebika.ai/help/guide/ai/ai-agentread 2026-10-18
Workflow Orchestration	Full
Automations chain triggers and actions with conditions, loops, delays, filters, random and round robin assignment, scripts and record operations, and a Call Agent action hands a step to an agent and routes the conversation to chosen members. Several agents work in parallel as a team, each running its task in the cloud while the user moves between them, and agents sit in the team directory grouped by department. Missions and reports created by automations assign follow up work to people. There is no supervisor agent that delegates to others on its own. Sourcebika.ai/help/reference/automation-action/call-agentread 2026-10-18
Knowledge Grounding & RAG	Partial
Each agent is limited to the data sources configured for it, which can be the space's databases, rich documents, automations and folders, so answers draw on the records and documents the team keeps in Bika. Intelligent search finds folders, databases, forms, missions and reports across the space. Bika offers no connector that indexes outside document stores, and chunking settings and source citations are not part of the agent setup. Grounding therefore depends on what a team has already brought into its space, which suits a business that runs its CRM and content calendar there and leaves a large document archive elsewhere out of reach. Sourcebika.ai/help/guide/ai/ai-agentread 2026-10-18
Human Oversight & Guardrails	Partial
In the Twitter agent example, the agent searches the web, drafts a post and publishes it only after the user approves it in chat, and agents can use only the data sources an administrator grants them. Automations route work to people through missions, record requests and approval style templates such as contractor reviews. Those approvals happen in conversation or through tasks a team sets up. There is no switch that makes an agent stop for sign off before a tool call, no guardrail layer on agent output and no queue where a reviewer clears agent actions. Sourcebika.ai/help/guide/ai/ai-agentread 2026-10-18
Security, Identity & Governance	Partial
Spaces organize people into members, roles, teams and guests, where members inherit the space's default permissions and guests see nothing until granted access, and every database, folder, automation, dashboard, document and agent can carry its own permissions by member, team or role. Agents can be kept private in a personal area. Team plans add organization level permissions and Business adds advanced permissions and audit logs. Sign in uses email, Google or WeChat, with no single sign on, and no security certification appears. Sourcebika.ai/help/guide/space/resource-permissionread 2026-10-18
Observability & Auditability	Partial
Business plans include audit logs, and each agent's full conversation history stays visible to members with access, so what an agent said and did in chat can be reviewed. Credit use is tracked against the space balance, with each model priced by a multiplier, and automation actions can be test run to confirm they work. Bika has no step by step trace of an agent run, no log of the tool calls an agent made and no run history view for automations, so investigating a bad result means reading the chat and checking the records it touched. Sourcebika.ai/pricingread 2026-10-18
Memory & State Persistence	Partial
Every agent keeps its conversation history, which members with access to the agent can reopen, continue and share by link inside the space or publicly, and long tasks keep running in the cloud when the user leaves the page. Durable state otherwise lives in the space's own databases and documents, which agents read and write through their skillsets, so a team that wants an agent to recall a customer's history keeps it in a database the agent can query. The agent does not build up its own notes about users or past cases between conversations. Sourcebika.ai/help/guide/ai/ai-agentread 2026-10-18
Deployment & Data Residency	Full
Besides the hosted service, Bika sells a self hosted edition installed on the customer's own servers for data sovereignty and compliance, requested through a license form that returns a deployment guide, an installation package and a 30 day trial key, then converted to a full license. Self hosted and white label editions are sold as enterprise licenses, with private deployment handled by the sales team. Hosting regions for the cloud service are not stated. Sourcebika.ai/help/guide/self-hosted/quick-startread 2026-10-18
Prebuilt Agents / Templates / Packs	Full
The template center offers 145 installable agentic apps that bundle agents, databases, automations, forms and dashboards, among them an Email Marketer with three day follow up sequences, a Twitter assistant, YouTube publishing automation, a GitHub issues creator, a lightweight CRM, invoice reminders, contractor management and a stock news reporter. Installing a template clones it into the space with its roles preset, ready to customize. Ready made agents such as AI Writer, Brand Designer and Office Docs Helper install on their own, and an AI wizard can generate a new agent template from a written requirement. Sourcebika.ai/sitemap/en/templates-sitemap.xmlread 2026-10-18
Triggers & Channel Coverage	Full
Automations start on a scheduler, an incoming webhook with its own URL, a new record or one that matches set criteria, a date field coming due, a submitted form, an inbound email to a connected IMAP account, a change in an HTTP response, or a manual run, with member joined and button click triggers listed as coming soon. Results go out by email, SMS, Slack, Telegram, WeCom, Feishu and DingTalk messages, reports and missions. People talk to agents through the in app chat and shared conversation links, and a message in Slack or Telegram does not start an agent conversation on its own. Sourcebika.ai/help/reference/automation-trigger/features-listread 2026-10-18
Model Flexibility & Routing	Full
Each agent picks its model, and the default Auto setting lets the system choose the best one and moves to newer models as they arrive. Models cost credits by a published multiplier, so a heavier model costs about twice a baseline one and a lighter model about half. Integrations connect OpenAI, Amazon Bedrock, Azure AI, Anthropic Claude, Google AI, DeepSeek, Alibaba Qwen and ByteDance Doubao, and automations offer separate text generation actions for OpenAI, DeepSeek and Bika's own model. Sourcebika.ai/help/guide/ai/ai-agentread 2026-10-18
APIs / SDKs / MCP Extensibility	Full
A REST OpenAPI with a published OpenAPI specification and an object oriented TypeScript SDK reaches spaces, nodes, databases, records and automations with bearer token authentication, with cURL, Python and TypeScript examples, alongside an APITable compatible API, a filter query language and cell value formats for developers. Agents extend through custom MCP servers over Streamable HTTP and the ToolSDK.ai catalog, automations run JavaScript or Python scripts, and webhooks bring outside systems in. API rate limits rise by plan, and Team plans add advanced OpenAPI access. Sourcebika.ai/help/guide/developer/openapiread 2026-10-18
Testing, Debugging & Optimization	Partial
Every automation action, including Call Agent and MCP server actions, has a Run test button that executes the step to confirm the flow works before it goes live, and interactive demos on reference pages show each feature's parameters. Agents are tuned by editing their prompt, data sources and skillsets and trying them in chat. There are no saved test sets, no scoring of agent answers and no way to compare one version of an agent against another, so quality checks stay manual. Sourcebika.ai/help/reference/automation-action/call-agentread 2026-10-18
Browser / Computer-use	Not documented
No skillset or automation action operates a browser or a desktop. Agents reach the web through search, HTTP requests and MCP servers, and an HTTP If Change trigger watches a URL's response for differences. Each of those reads content, and none clicks, types or signs in on a page, so a site with no API is reachable only for reading. A team could attach a browser automation server from the ToolSDK.ai catalog on its own, but Bika neither ships nor names one, and none of its templates use a browser. Sourcebika.ai/help/reference/automation-trigger/features-listread 2026-10-18

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

Pricing

Plus $19 per seat a month, or $16 a month billed yearly

Per seat monthly or yearly subscription with an AI credit allowance per seat; credits pay for AI inference at a per model multiplier

Free tierTrial available

Included quota

Free: 3,000 daily credits, 5 GB, 10,000 records per database, 200 automation runs, 5 seats. Plus: 50,000 credits per seat, 20 GB, 50,000 records per database, 30,000 runs. Team: 100,000 credits per seat, 200 GB, 500,000 records, 60,000 runs. Business: 500,000 credits per seat, 800 GB, 1,500,000 records, 100,000 runs. Paid plans currently double the credits as a limited offer.

What is public

All four cloud plan prices with monthly and yearly rates, credit, storage, record and automation allowances, and the credit rules.

Billing mechanics

Self serve per seat billing, monthly or yearly, with credits pooled into a shared space balance that refreshes each billing date.

Cost watchouts

Powerful models burn credits at higher multipliers, the current doubled credit allowance is a limited offer, and audit logs need the Business plan at $199 per seat a month.

Variable cost rationale

Seat prices are fixed, but agent conversations and AI actions draw space credits at a per model multiplier, so heavier models or busy agents can use up the monthly allowance, and unused credits do not roll over.

Additional watchouts

Business is far above Team per seat, so audit logs carry a steep step up.

Overage / add-ons

Subscription credits refresh on the billing date and do not roll over; space credits also include daily and gift credits, and referral credits can be transferred into a paid space. A price for buying extra credits is not set out.

Sales call required

No, self serve available

Free / trial

Free plan for up to 5 seats with 3,000 daily AI credits and 200 automation runs; 30 day trial key for the self hosted edition

Lowest paid plan

Plus at $19 per seat a month, or $16 billed yearly: 50,000 AI credits per seat, 20 GB storage, 30,000 automation runs

Commercial notes

Team adds organization permissions, email support and advanced API access; Business adds audit logs, full integrations and advanced permissions. Self hosted, white label and private deployment are quoted.

Key ambiguities

How many agent tasks a seat's credits cover depends on which models agents use and how long their inputs and outputs run.

Missing data

Self hosted and enterprise license pricing and the price of extra credits.

Agentic Index verified 2026-10-18

Official pricing page Docs

Data confidence: medium