agentboards.org
Compare/Devin vs Ona

DevinvsOna

Generated from the two spec rows. Green marks the better value where a spec has a clear direction. Everything else is just different.

Devin
Cognition · Autonomous SWE
#139MCP
Panel
5.3
3 spec wins
Reliability
5.3
Usefulness
5.8
Cost
4.3
Longevity
5.5

“Replaced ACUs with credits in April 2026, so now the meter runs in a unit you already understand.”

Ona
Ona (OpenAI) · Autonomous SWE
#31MCP
Panel
6.6
1 spec wins
Reliability
6.7
Usefulness
7.5
Cost
5.0
Longevity
7.3

“Automations fan a fleet of agents across your codebase on a schedule, so now the on-call rotation includes the robots.”

Spec by spec

SpecDevinOna
Architecture
CategoryAutonomous SWEAutonomous SWE
Runscloud, sandbox, localcloud, sandbox
Platformsweb, macos, linux, windowsweb, macos, linux, windows
Context windownot documentednot documented
Protocols
MCP clientYesYes
MCP serverYesNo
Capabilities
Runs terminal commandsYesYes
Multi-file editsYesYes
Git operationsYesYes
Browser controlYesCloud sessions ship a first-party interactive Browser tool alongside the shell and IDE, with saved browser profiles for authenticated sites (https://docs.devin.ai/work-with-devin/browser-auth). No
Sandboxed executionYesCloud sessions run on a dedicated Devin machine built from environment blueprints and snapshots, and the Devin CLI adds OS-level isolation locally (https://docs.devin.ai/cli/sandbox). Yes
Multi-agent orchestrationYesYesAutomations run fleets of agents triggered by pull requests, schedules or webhooks.
Headless / CI modeYesYes
Models
BackboneUndisclosed (Cognition-managed models)OpenAI Codex
Bring your own modelNoA cross-provider model picker (`--model` / `/model` / config `agent.model`) selects between Anthropic, OpenAI, Google, Cognition and open-source models, but every one of them is served by Cognition: the CLI model reference documents no API key, base URL, gateway, Bedrock, Vertex or Azure deployment of your own, matching pricing.byok = false. No
Local modelsNoNo base URL, gateway or local endpoint setting exists in the Devin CLI config reference (https://docs.devin.ai/cli/reference/configuration/config-file). No
Cost
Pricing modelmixedmixed
Starts at$20/mo$20/mo
Free tierYesNo
Bring your own keyNoUsage is billed in Cognition ACUs and no LLM provider key can be supplied; Devin Outposts (https://docs.devin.ai/cloud/outposts/overview) only moves session compute onto your own machines. YesA Codex subscription can be connected to a Core plan.
Openness
Open sourceNoNo
Licenseproprietaryproprietary
GitHub starsn/an/a

Which one would each critic pick

CriticDevinOnaPick
El Juez——not enough reviews
El Amigo6.36.8Ona — Pick Ona when you want to hand over a ticket and get a pull request without opening a branch; pick Devin if you want the same shape with a longer track record.
El Crítico5.36.3Ona — The recommended backbone narrowed to one lab's agent and the earlier alternative is documented as deprecated for Enterprise, so model diversity here is ending, not expanding.
El Profesor5.07.3Ona — Verification has a real substrate: the agent works inside an environment carrying your toolchain, permissions and network access, so a claim can be executed rather than argued.
La Inversora6.87.8Ona — This was Gitpod, renamed in 2025, and it is now part of the OpenAI Platform, so the cap-table question every buyer asks has already been answered.
La Jefa5.87.3Ona — Core starts at $20 a month with 80 to 2,200 compute units and top-ups from $10 per 40, and Enterprise runs inside our VPC with kernel-level policy and audit trails.
El Hacker2.54.5Ona — Closed source and cloud-only, but it speaks MCP and a Codex subscription can be attached to a Core plan, which is the one string I get to hold.

Picks are derived from each critic's own scores. Humans vote on matchups on the duels page.

Want a third column? The compare tool handles any two agents. Three-way comparisons are on the roadmap once the spec rows are all verified.