Tabnine
tabnine.comSummary
Tabnine is an AI code assistant designed to accelerate and simplify software development. It provides AI-powered code completions, chat, and agents for tasks like code generation, explanation, testing, and documentation. Tabnine emphasizes privacy, personalization, and protection, offering various deployment options including on-premises and air-gapped solutions for enterprises.
Features8/13
See allMust Have
5 of 5
Conversational AI
API Access
Safety & Alignment Framework
Fine-Tuning & Custom Models
Enterprise Solutions
Other
3 of 8
Code Generation
Security & Red Teaming
Synthetic Media Provenance
Image Generation
Multimodal AI
Research & Publications
Threat Intelligence Reporting
Global Affairs & Policy
PricingSubscription
See allDev Preview
- AI code completions for current line and multiple lines for full-function implementation
- AI-powered chat using leading LLMs from Anthropic, OpenAI, and Tabnine
- Foundational AI agents
- Support for all major IDEs
Dev
- AI chat in the IDE supports every step in the SDLC
- AI agents autonomously generate code, tests, docs, fixes, and more
- Powered by best-in-class AI models from Tabnine and leading LLM providers
- Basic personalization through Local IDE context awareness
- Integration with Atlassian Jira Cloud to inform AI responses and generation
- Enterprise-grade security, safety, and privacy
- Access to Tabnine models exclusively trained on permissively licensed code
- Ticket-based support available during business hours
Enterprise
- All Dev features, plus
- Advanced AI agents: test case agent, Jira implementation agent, code review agent
- Fully private deployment on SaaS or self-hosted (VPC, on-premises with the option to be fully air-gapped)
- Advanced Context Engine (Unlimited codebase connections for Bitbucket, GitHub, and Gitlab)
- Integration with Atlassian Jira Cloud and Data Center to inform AI responses and generation
- Customized AI Code validation rules that enforce your standards for code quality
- Model flexibility: use models from third party providers, open source models, and your own internally developed models
- Code provenance and attribution identifies the source and license of AI generated code from any LLM
- Admin censorship controls to block code injections that match non-compliant licenses
- IP indemnification
- Priority ticket-based support during business hours
- Training on AI-enabled software development for your entire team
- Advanced admin tools for user and policy management
- Governance for model selection and repository connection permissions
- Admin tools for tracking usage and user productivity
- SSO integration for ease of administration for private deployments
Rationale
Tabnine is a strong match for the OpenAI Platform concept. It offers conversational AI through its chat interface, API access for integration into IDEs, and a robust safety and alignment framework with features like IP indemnification, permissively licensed training data, and provenance/attribution. The platform also provides fine-tuning capabilities for custom models based on a user's codebase and offers dedicated enterprise solutions with private deployments (SaaS, VPC, on-premises, air-gapped) and advanced administration tools. Furthermore, it explicitly offers code generation, code explanation, and code fixing, which aligns with the 'code-generation' feature. The 'security-and-red-teaming' feature is covered by its focus on security, privacy, and protection from risk, including checks against publicly viewable repositories. The 'synthetic-media-provenance' is directly addressed by their 'Provenance and Attribution' feature for AI-generated code.
