Sourcegraph
Also known as: Deep Search, Code Search, Cody, Amp
Code intelligence platform with Cody AI assistant, Deep Search, and Amp for enterprise developer productivity and code understanding.
Sourcegraph is a code intelligence platform built to help both engineers and AI agents understand, oversee, and safely change very large, complex codebases. Founded in 2013, it is used by major engineering organizations and is best known for its code search: the ability to search across every repository, branch, and code host with exact, exhaustive results, powered by a code graph that maps the relationships between symbols, references, and definitions throughout a codebase.
That code graph anchors a wider set of capabilities. Deep Search lets a developer ask a complex question in natural language and get a grounded answer with citations from an agent that explores the repositories on its own. Batch Changes rolls out the same fix, refactor, or migration across many repositories at once with control at scale. And tools for code insights, monitoring, and living documentation track how a codebase evolves over time and flag risky changes as they happen.
In the agentic era, Sourcegraph has positioned its code graph as the context layer that makes AI coding agents reliable. Because an agent working on one task usually sees only fragments of an enterprise codebase, it can miss changes or introduce inconsistencies; Sourcegraph indexes all of an organization's repositories and, through an MCP server, feeds agents complete, precise context so they produce better results with fewer retries and lower inference cost.
Sourcegraph also ships its own AI coding tools. Cody is its assistant for writing, fixing, and understanding code inside IDEs, drawing on full-codebase context and letting developers choose among frontier models from Anthropic, OpenAI, and Google; as of 2025 Cody is focused on enterprises. For individual developers, Sourcegraph offers Amp, an agentic coding tool with multi-step task automation that taps the same code graph for context.
Built for the largest monorepo and multi-repository architectures, Sourcegraph supports cloud, self-hosted, and air-gapped deployments, keeps LLM inference data out of retention, and meets enterprise requirements with SOC 2 and ISO 27001 compliance, single sign-on, and role-based access. Its throughline is giving humans and agents complete, trustworthy context at the scale of Big Code.
Vendor details
Canonical URL
https://sourcegraph.com/
Category
Coding agent
Company status
independent
Use cases & customers
Target customers
Deployment options
In practice
Your AI agents keep making inconsistent changes because each one only sees a slice of a sprawling enterprise codebase. Sourcegraph indexes every repository and feeds agents complete code-graph context through its MCP server, so their output is reliable.
You need to apply the same security fix across hundreds of repositories without missing any. Sourcegraph's Batch Changes rolls the change out everywhere at once, with control to review and track it at scale.
A new engineer needs to understand how a function is used across the whole system. Sourcegraph's code search and Deep Search return exact references and a natural-language answer with citations, instead of grepping through hundreds of files.
Agentic Index coverage score
8.0 / 14 capabilities · 57%
| Integrations & Tool CallingAgentic AI Index Expansion Evidence Report | Full |
|---|---|
| Workflow OrchestrationAgentic AI Index Expansion Evidence Report | Partial |
| Knowledge Grounding & RAGAgentic AI Index Expansion Evidence Report | Full |
| Human Oversight & GuardrailsAgentic AI Index Expansion Evidence Report | Unable to verify |
| Security, Identity & GovernanceAgentic AI Index Expansion Evidence Report | Full |
| Observability & AuditabilityAgentic AI Index Expansion Evidence Report | Full |
| Memory & State PersistenceAgentic AI Index Expansion Evidence Report | Unable to verify |
| Deployment & Data ResidencyAgentic AI Index Expansion Evidence Report | Full |
| Prebuilt Agents, Templates & PacksAgentic AI Index Expansion Evidence Report | Unable to verify |
| Triggers & Channel CoverageAgentic AI Index Expansion Evidence Report | Partial |
| Model Flexibility & RoutingAgentic AI Index Expansion Evidence Report | Partial |
| APIs, SDKs & MCP ExtensibilityAgentic AI Index Expansion Evidence Report | Full |
| Testing, Debugging & OptimizationAgentic AI Index Expansion Evidence Report | Partial |
| Browser & Computer UseAgentic AI Index Expansion Evidence Report | Unable to verify |
The Agentic Index coverage score grades every vendor Full, Partial or Unable to verify 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
Recent platform changes
Sourcegraph is reinforcing technical safeguards to strictly enforce geographic access restrictions across its products. This update aligns access controls with applicable laws and the requirements of underlying AI model providers.
Bears on: Deployment / data residency
View sourceSourcegraph released Code Finder in Beta through its Model Context Protocol (MCP) server, an agentic tool that helps AI coding agents rapidly locate relevant files and documentation within single repositories. Additionally, Sourcegraph made its redesigned Compare page generally available, introducing a file tree view and full code navigation for analyzing complex diffs.
Bears on: MCP / tool calling / API
View sourceSourcegraph launched Agentic Batch Changes in public beta: an agent harness for applying code changes across thousands of repositories or the largest monorepos. Built on Batch Changes and Deep Search, the agent plans a change, chooses between writing deterministic scripts or delegating per-repo work to Claude Code or Codex, executes across all targets, and reacts to downstream events like CI failures and merge conflicts until changes are ready to merge. Beta access is by request.
Bears on: Agent capability
View sourcePricing
Cody Enterprise $59/user/mo (annual) · platform ~$16K/yr+
seats
Included quota
Sourcegraph Free: basic code search across public repos + limited private support. Cody Enterprise ($59/user/mo, annual): full platform - code search + AI chat/autocomplete/inline edits, cross-repository context (indexes ALL org repos, 10+ at once), Context Filters, admin-chosen LLM (Claude Sonnet/Opus, GPT-4o, o1, Gemini), single-tenant deployment, no-training guarantee, SOC 2, self-host/BYOC/air-gapped. Enterprise platform adds code insights, batch changes, and advanced security at scale. Amp (separate company since Dec 2025): agentic CLI + VS Code coding with parallel subagents, Oracle/Smart modes, free credits + PAYG.
What is public
Sourcegraph (sourcegraph.com - enterprise code-intelligence platform: code search, navigation, batch changes, insights + AI) is enterprise-focused and quote-negotiated. Vendr data: annual contracts run ~$15-16K (small teams, 10-25 users) to $250K+ (hundreds of devs), roughly $50-$200+/user/mo, with self-hosted minimums ~$50K-$75K/yr; the pricing page now frames Sourcegraph Enterprise as starting ~$16K with included AI-feature credits. A free tier covers basic public-repo code search. MAJOR 2025 change: Cody Free and Pro were DISCONTINUED (signups ended June 25, 2025; terminated July 23, 2025) - Cody is now ENTERPRISE-ONLY at $59/user/mo (annual), and individual developers were moved to the agentic product Amp (free credits + pay-as-you-go). Amp was then SPUN OFF into a separate company in December 2025.
Billing mechanics
Per-seat, annually negotiated for the platform + Cody Enterprise ($59/user/mo list, often discounted to ~$20-$30/user/mo for 25+ seats with competitive quotes); priced by active-user count, feature tier (code search vs. insights/batch changes), and deployment model (cloud vs. self-hosted/single-tenant/BYOC). Amp - now a separate company (spun off Dec 2025) - is pay-as-you-go with zero-markup model pricing (you pay model API cost) plus a free-credit on-ramp ($10 on signup) and an enterprise option. Multi-year commitments unlock better per-user rates.
Cost watchouts
Cody is enterprise-ONLY since July 2025 - no individual/small-team option (Cody Free/Pro were sunset), and $59/user/mo is ~3-6x GitHub Copilot for individuals; 'active user' definitions matter (clarify in contract to avoid overage); self-hosted lowers per-user license but adds infrastructure/maintenance cost; the removal of Free/Pro eliminated a low-friction evaluation path; Amp's PAYG model usage is uncapped (set your own limit)
Variable cost rationale
Platform/Cody is per-seat (predictable once seat count is set); Amp is genuinely usage-based (zero-markup model spend scales with agent runs); overall exposure depends on which product you adopt and your developer headcount
Additional watchouts
No individual/small-team Cody option anymore (enterprise-only, $59/user/mo, annual commitment) - solo devs should use Amp (now a separate company) or competitors; pricing is opaque/negotiated; Cody is more chat/autocomplete-focused than fully autonomous agents (Amp/Cursor/Claude Code lead there); confirm 'active user' definition and map features to tiers to avoid overpaying
Overage / add-ons
Platform/Cody: per-seat, so cost scales with active-user count; exceeding contracted seats or adding modules (insights/batch changes) is renegotiated. Amp: pay-as-you-go at zero-markup model rates - spend scales directly with agent usage after the free credits.
Sales call required
Mixed (some tiers require a call)
Free / trial
Team cloud trial available
Lowest paid plan
Enterprise from ~$16K/yr
Commercial notes
Founded 2013, SF (Stanford grads Quinn Slack & Beyang Liu); ~800k developers, indexes 50B+ lines of code; early customers Uber/Dropbox/Lyft; core moat is cross-repository code-graph context at enterprise monorepo scale (beats single-repo tools like Copilot/Cursor for 'how is this used across our codebase'); strategically split in 2025 into Cody Enterprise (large orgs) + Amp (agentic), with Amp spun off into a separate company Dec 2025; competes with GitHub Copilot Enterprise ($39), Cursor Business ($40), Tabnine Enterprise ($39-$59)
Key ambiguities
Pricing is negotiated and highly variable ($50-$200+/user/mo observed; small teams ~$15-16K/yr to enterprise $250K+); some 2026 listings still cite the OLD Free/$9 Pro Cody tiers - those are discontinued (current Cody is enterprise-only $59/user/mo); Amp (now a separate company) - its exact PAYG/enterprise rates beyond 'zero-markup + free credits' aren't fully public
Cancellation / refund
Annual contracts standard for the platform + Cody Enterprise (multi-year unlocks discounts); a free public-repo code-search tier and a ~30-day team trial path exist; Amp offers free credits ($10 signup) + pay-as-you-go with no lock-in; Cody Pro/Enterprise-Starter users were offered $40 Amp credits during the 2025 transition
Support SLA / resale
Enterprise: dedicated support, SOC 2, no-training-on-your-code guarantee, single-tenant/self-hosted/BYOC/air-gapped deployment, admin LLM selection; integrates across VS Code, JetBrains, Neovim and code hosts (GitHub/GitLab/Bitbucket/Perforce); Amp adds enterprise-grade security for agentic workflows
Missing data
Platform/Cody pricing is negotiated (wide $50-$200+/user/mo range; ~$15-16K/yr small-team to $250K+ enterprise); Amp's full PAYG/enterprise rate card isn't public. Seed '~$16K/yr Enterprise; self-hosted/single-tenant available' is accurate for the small-team platform entry (Vendr ~$15K; pricing page ~$16K), but note the MAJOR 2025 changes: Cody Free/Pro discontinued (June/July 2025), Cody now enterprise-only at $59/user/mo, Amp launched as the new individual-developer product, and Amp was SPUN OFF into a separate company in Dec 2025.
Related vendors
- Cognition — Maker of Devin, an autonomous AI software engineer
- 10Web — Agentic website platform whose team of ten specialized AI agents…
- Ageiro — Early stage enterprise platform that turns business goals into…
- Aider — Open-source, model-agnostic terminal coding agent that works…
- Anthropic Claude Code — Anthropic's terminal and IDE-native coding agent with sub-agent…
- AppFactor — Agentic orchestration platform that autonomously maintains and…
Alternatives to Sourcegraph
The closest documented capability profiles to Sourcegraph among coding agents tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.
- Augment Code10.5 / 14Adds documented Memory & State Persistence and Prebuilt Agents, Templates & Packs
- Greptile8.5 / 14Adds documented Human Oversight & Guardrails and Browser & Computer Use
- Continue10.5 / 14Adds documented Human Oversight & Guardrails and Memory & State Persistence, among others
- Poolside10.5 / 14Adds documented Human Oversight & Guardrails and Prebuilt Agents, Templates & Packs, among others
- Qodo8.5 / 14Adds documented Human Oversight & Guardrails and Memory & State Persistence, among others
- Refact.ai8.5 / 14Adds documented Human Oversight & Guardrails and Browser & Computer Use
Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded