To learn more about large language models, see Introduction to LLMs.
Compare models
| Model | Speed | Reasoning depth | Best use |
|---|---|---|---|
| Haiku 4.5 | Fastest | Light | Quick edits and iteration |
| Sonnet 4.6 (default) | Balanced | Deep | Most everyday development work |
| Opus 4.6 | Slower | Deepest | Advanced reasoning, legacy codebases, prompt-directed deep thinking |
| Opus 4.7 | Slower | Deepest | The most complex builds, large projects, output quality at scale |
Not all models are available on the free plan. To access all models, upgrade to a paid plan.
Haiku 4.5
Fast and token-efficient for simple tasks and rapid iteration.Best for:
- Quick UI or styling changes
- Content edits and translations
- Test or seed data
- Straightforward static pages
- Tasks where the solution is already clear
Sonnet 4.6
The default model in Bolt. Strong reasoning depth with speed and efficiency built for all-day use.Sonnet 4.6 is the model most users should start with, so it’s selected by default for new projects. It offers a balance of speed, cost, and reasoning depth for everyday building tasks. It performs well across planning, coding, debugging, and iterative development. It can hold a complex plan across multiple steps and execute without losing the thread.It can handle multi-file refactors, end-to-end feature builds, and coordinated front-and-backend changes smoothly. For most development work, Sonnet 4.6 delivers excellent results while staying fast and token-efficient enough to use throughout the day.Best for:
- Scaffolding new apps and features
- Backend logic and database work
- Multi-step feature builds spanning front and backend
- Complex refactors touching many files
- General-purpose development where reliability and speed both matter
Opus 4.6
A highly capable model with deep reasoning and high sensitivity to your instructions.Opus 4.6 is tuned to respond more directly to how you prompt it. If you tell it to think carefully about a specific problem, it invests deeper reasoning in that area in a way other models may not. This makes it especially well-suited for navigating large, existing codebases and reasoning through messy or legacy code where understanding tangled dependencies and older patterns matters. For new projects or tasks where Sonnet 4.6 already performs well, Opus 4.6 may not show a meaningful difference. Its strength emerges in the hardest problems where precise, directed reasoning makes a real impact.Best for:
- Very complex architectural and system-wide decisions
- Brownfield development in large, tangled codebases
- Tasks where you want fine control over reasoning depth through your prompts
- High-stakes work where the extra reasoning investment is justified
Opus 4.7
The most capable model available in Bolt, with stronger reasoning across complex, long-running tasks.Opus 4.7 shows improved performance on tasks that require analyzing and synthesizing large amounts of information. It builds more complete features with fewer gaps and errors, even on the first try.Opus 4.7 is also better at monitoring its own work. As it builds, it can assess and improve the quality of what it creates. It also handles long tasks more reliably: it can sustain focus across extended builds without losing track of its original plan.For new projects or tasks where Opus 4.6 already performs well, Opus 4.7 may not show a dramatic difference. Its strengths are most noticeable on long, complex builds and on projects where getting things right the first time saves significant rework.Opus 4.7 uses more tokens per task than Opus 4.6. The additional usage goes toward more complete code, more thorough setups, and deeper reasoning rather than wasted effort.Best for:
- The most complex, multi-step builds where output quality matters most
- Large projects that require reasoning across many files and dependencies
- Tasks that involve data analysis or synthesis
- Work where self-correction and output quality reduce the need for manual review
Switch models
You can switch models in two places: from the Bolt homepage when starting a build, or inside a project while working. For each project, Bolt remembers the model you select, and it automatically uses that model again the next time you open the project. If you want all new projects to use a specific model, you can set a default model in your personal settings. Bolt then automatically selects that model when you start a new project.Switch models on the homepage
Switch models in a project
- In the bottom-left corner of the chatbox, click the current model name.
- Select the model you want to use.
Use claude.md files
Bolt supports importingclaude.md files directly into projects. When a claude.md file is present, it’s detected and used automatically, with no additional setup required.
To upload a file:
- Open your project.
- Click the plus icon in the chatbox.
- Click Attach file, and choose your file.
claude.md file is especially helpful for users who rely on Project Knowledge settings. Instead of adding content manually through the UI, you manage project context in Markdown files. This makes it easier to include links, organize information across multiple files, and maintain project knowledge in a more flexible and scalable way.
The
claude.md file acts as the entry point for agent instructions. While it can reference other files, such as text or additional Markdown files, the primary instructions must live in claude.md itself. The file name is required and can’t be changed. Agents only look for instructions starting from claude.md, then follow any links or references you include from there.How Bolt handles context
All Claude models support large context windows. However, to keep performance smooth, Bolt limits the active history to recent messages. This helps reduce token usage and keeps building responsive. If a detail from earlier becomes important again, briefly restate it to bring it back into context. For any ongoing context or instructions that are important, add them to either your Project Knowledge, Account Knowledge, or, if applicable, Teams Knowledge.Projects using v1 Agent (legacy)
As of April 13, 2026, you can no longer select v1 Agent (legacy) when starting a new project. The v1 agent will be retired on August 3, 2026, and any remaining projects that use it will be automatically switched to the Claude Agent default model.