Cosine review, pricing and verdict

Coding agent for CLI, desktop and cloud, with air-gapped Lumen models.

Best for
Engineering leads who want many coding agents running in parallel in the cloud
Not for
Solo developers who want in-editor autocomplete
  • Coding

Should you use Cosine?

Cosine suits engineering teams that need an agent for legacy languages or a closed network and want to swap models. It is a poor pick for hobbyists who want autocomplete, and its Lumen Sovereign model is not due before 2027, so judge it on the agent it ships today, not the roadmap.

A coding agent backed by the company's own models and a deployment mode that keeps code, prompts and outputs inside the customer's boundary.

Also for

  • Non-engineering teams using Desktop for research and spreadsheet work
  • Engineering teams that maintain COBOL, ABAP, Fortran or Verilog systems
  • Platform and security teams at banks, defense contractors and public bodies that need air-gapped AI

Also not for

  • Teams that need a fixed price per task

Main trade-off

The flagship sovereign model is not due before 2027 and has no published results.

Cosine pricing

Top plan
$999/mo
Billing model
Subscription

Credits cover agent work, model calls and cloud execution, so a task's cost depends on its size, the model chosen and the runtime rather than on tokens. Every tier can buy top-up credits at any time. Enterprise and private deployments, including air-gapped ones, are scoped and priced with Cosine's sales team.

What Cosine does on its own

Scope of autonomy
Semi-autonomous
Human approval boundary
Human in the loop
Maturity
Generally available

Safety controls

  • Runs sandboxedSupported
  • Audit loggingSupported
  • Human in the loopSupported
  • Output filteringNot supported

What it can operate

  • Code execution
  • File system
  • Shell commands

Where Cosine runs

Deployment
Cloud · Local · Hybrid
Where it runs
Web · Desktop · CLI
Memory
Long term

Models it runs on

  • Lumen Outpost
  • Lumen Scout
  • OpenAI models
  • Anthropic Claude models
  • Google Gemini
  • Kimi
  • MiniMax
  • GLM

Systems it connects to

SystemType
GitHubNative
SlackNative

Languages it supports

  • Python
  • JavaScript
  • Rust
  • C
  • C++
  • C#
  • Java
  • Scala
  • COBOL
  • ABAP
  • Fortran
  • Verilog

What Cosine does

Cosine is a London AI lab that sells a coding agent for the command line, desktop and cloud, built on its own Lumen models and a menu of third-party ones. The agent plans, edits and tests code with checkpoints and approval-gated actions, and Cosine offers air-gapped deployment for regulated buyers.

  • Checkpointing

    Every agent turn in the CLI is a lightweight git commit, so any change can be undone instantly.

  • Swarm subagents

    In Swarm mode an orchestrator spawns specialised child agents that work on parts of a task at the same time.

  • Parallel cloud agents

    Twenty-plus agents work at once on isolated worktrees, each in a MicroVM that Cosine says boots in about 125 ms.

  • Approval-gated actions

    Reads run automatically, while writes such as pushing code, sending a Slack message or filing a pull request comment wait for one confirmation.

Show all 8 features
  • One model menu

    Lumen Outpost and Lumen Scout sit beside OpenAI, Anthropic, Google, Kimi, MiniMax and GLM models, with a cost multiplier shown for each in Desktop.

  • Legacy-language focus

    Its niche-languages page names COBOL, ABAP, Fortran and Verilog, with repository mapping, test scaffolding and documentation for old code.

  • Air-gapped deployment

    Runs where source code, prompts and outputs cannot leave the boundary, with audit visibility into agent actions, tool use and review status.

  • Cosine Desktop

    A Windows app that applies the agent to spreadsheets, vendor research and customer-feedback review for non-engineers.

Strengths and limits

Strengths

  • The air-gapped option is documented: Cosine's air-gapped page promises no data egress and says every change stays visible, testable and subject to customer review.
  • The base cost can be read without a sales call: three tiers and per-million add-on credit rates are published on its pricing page.
  • Cosine publishes its own benchmark suite for Lumen Outpost (Niche-Bench, Vibe-Bench, Slop-Bench and a cost-per-successful-task measure), though the figures are the vendor's.

Limits

  • Lumen Sovereign does not exist yet: the homepage dates its first release from Q1 2027, later than the late-2026 target in Cosine's June 2026 blog, so its sovereignty claims cannot be tested.
  • Independent doubt exists: the Financial Times, as summarised by AI Weekly on 11 August 2026, questioned whether a lab of roughly 30 people has the talent and compute to train a frontier model.
  • Hokai found no independent benchmark score for Lumen Outpost; the figures on Cosine's page are its own.
  • Credits are hard to forecast: the pricing page says use depends on task size, model choice and runtime, so Hokai scores value for money at 6.0 despite the published tiers.

How Hokai rates it

  • Features8.0 / 10
  • Value for money6.0 / 10

Cosine alternatives

AgentWhere it runsPriceMaturityRating
Cosine This pageCloud · Your machine$19/moAvailable3.5
Claude CodePick Claude Code when you want Anthropic's own agent across terminal, IDE, desktop and web; pick Cosine when you need a swappable model menu or an air-gapped deployment.--------
DevinPick Devin when you want a hosted autonomous engineer with a free plan to start; pick Cosine when model choice or air-gapped deployment decides the purchase.--------
OpenHandsPick OpenHands when you want an open-source, model-agnostic coding agent that you run yourself; pick Cosine when you want a managed platform with parallel cloud agents and an air-gapped option.Cloud · Your machine · APIFree planAvailable--

Cosine facts

Works offline
Yes
Minimum skill level
Intermediate
Budget fit
Enterprise
Open source
No
Ships
Monthly
Last major update
Maker founded
2022
Headquarters
London, United Kingdom
Total funding
About $8M from Lakestar, SOMA Capital and Gaingels (tech.eu, April 2026)
Team size
About 30 (Financial Times, as summarised by AI Weekly, August 2026)

What changed

  • Release Cosine joins Leonardo's SME Collaboration Programme source
  • Integration Partnership with Shift on sovereign AI for operational resilience source
  • Integration Partnership with Mimic brings Cosine into cybersecurity source

Questions about Cosine

What does Cosine cost per month?

Cosine lists Starter at $19 a month with 4M credits, Team at $199 with 47M credits and Enterprise at $999 with 240M credits. Add-on credits cost $6.50 per 1M on Starter, $5.00 on Team and $4.50 on Enterprise. Enterprise and private deployments are priced with sales (pricing page, read 15 November 2026).

Does Cosine have a free plan?

No free plan or trial appears on the pricing page, so the Starter tier is the cheapest way in. Some third-party directories still list an older free allowance, which the current page does not show.

What is Lumen Sovereign, and when does it arrive?

Lumen Sovereign is Cosine's planned frontier model, trained on the Isambard-AI supercomputer with UK government backing, in 1T and 200B parameter variants. Its homepage gives Q1 2027, which is later than the late-2026 target in the 8 June 2026 coalition post, and no results are published, so nothing can be tested yet.

How does Cosine compare with Claude Code?

Claude Code is Anthropic's own agent for the terminal, IDEs, desktop, web and mobile, tied to Anthropic's models. Cosine ships a menu that mixes its Lumen models with models from OpenAI, Anthropic, Google and others, and offers air-gapped deployment. Hokai found no independent head-to-head test of the two.

Does Cosine work on Windows?

Yes. The CLI installs with winget add Cosine.CLI, and Cosine Desktop has a Windows download aimed at tasks such as spreadsheet reports and vendor research. Cosine Cloud runs each agent session in a MicroVM on Cosine's workers or on your own SSH targets, so heavy work need not run on the laptop.

About Cosine

Cosine is a London AI lab founded in 2022 by Alistair Pullen (CEO) with co-founders Yang Li and Sam Stenner, a Y Combinator graduate that tech.eu said in April 2026 had raised $8 million from Lakestar, SOMA Capital and Gaingels. In April 2026 the UK government chose it as one of the first partners in its £500 million Sovereign AI programme, with 500,000 GPU hours on the Isambard-AI supercomputer, and the FT, relayed by AI Weekly in August, described a team of roughly 30. Its homepage now leads with sovereign models, "We train AI models for organisations that need frontier capability inside secure environments", while its docs describe Cosine as an AI coding agent for software engineering work, from quick fixes to complex migrations.

Read the full background

The agent works across three surfaces. The CLI makes every agent turn a lightweight git commit, so a change can be undone at once, and it keeps a live to-do list, project memory and, in Swarm mode, an orchestrator that spawns child agents. Cosine Cloud runs 20 or more agents in parallel on isolated worktrees, each session in a hardware-isolated MicroVM; reads run on their own and writes wait for a confirmation. Cosine Desktop applies the same agent to spreadsheets and research for non-engineers. Its model menu mixes the company's Lumen Outpost and Lumen Scout with models from OpenAI, Anthropic, Google, Kimi, MiniMax and GLM.

It fits teams that maintain systems in COBOL, ABAP, Fortran or Verilog, which its niche-languages page names as a focus, and regulated buyers that cannot send code out: its air-gapped page promises no data egress, a choice of Lumen, open-weight or customer-trained models, and a record of agent actions. Lumen Sovereign is the planned frontier model for that market, in 1T and 200B parameter variants; a coalition of UK banks, defense and telecom firms signed memoranda to co-design it in June 2026, and the homepage now gives Q1 2027 for the first release, against a late-2026 target in that announcement.

Skip it if you want in-editor autocomplete, for which Cursor is the usual choice, or if you want one vendor's agent across every surface: Claude Code is Anthropic's, Devin is Cognition's hosted engineer and OpenHands is an open-source, model-agnostic option you run yourself. The AI coding agents list shows the rest. Lumen Sovereign has no published results to test, so judge Cosine on the agent it ships today. Pick it only if legacy languages or a closed network decide the purchase; for everything else, start with Claude Code or OpenHands.

Sources and guides

More AI Agents on Hokai

  • OrbiTurns a labelled GitHub Issue into a reviewed, merged PR and a release
  • LetterCoding agent that turns site and market signals into pull requests
  • Muse CodeMeta's multi-agent CLI that plans, codes, and validates changes across repos.
  • RopilotA desktop agent that writes Luau into Roblox Studio and playtests it
  • OpenAI CodexClones your repo, writes the code, opens the pull request. No keyboard required.
  • PromptlessAI teammate that drafts doc updates as pull requests when you ship.