Platform extensibility
Which AI agent platforms let you choose or bring your own model?
This is the largest documentation gap in the index. Of 946 vendors, only 305 document full model flexibility: working across several foundation models, routing tasks between models, or letting the buyer bring their own model and their own provider keys. 135 document partial coverage and 506, more than half, document none at all.
Every vendor in the index is assessed against the same 14 point taxonomy from public documentation, and no vendor pays for placement. Counts on this page were measured across all 946 public vendors on September 30, 2026.
How the 946 vendors split
No public evidence means the reviewed sources did not document the capability. On this index that is a statement about the evidence, not proof that the capability is absent. See methodology.
What counts as full coverage
This axis measures who chooses the model, not how many models are involved. Full coverage means model choice is a documented buyer facing surface: a provider list the customer or admin selects from, a routing policy the customer controls, or bring your own model and bring your own key (BYOK). Partial means the vendor discloses that it runs or routes across more than one provider but keeps the choice itself. None means a single provider or nothing disclosed; a named single provider is the clearest None on the axis, since the absence is visible rather than inferred. Shipping an MCP server so that your assistant can reach the vendor is the opposite direction and is credited on extensibility, not here. Roadmap statements do not count.
How to read these numbers
Read this axis as a question about lock in rather than about capability. Multi agent platforms at 69 percent, agent builders at 68 percent and coding at 68 percent treat model choice as a feature because their buyers ask about cost and about the next release. Healthcare at 4 percent and GTM at 7 percent are the extremes, and the reason is that those products are sold as outcomes: nobody buys a prior authorization agent and then asks which model wrote the letter. That is a defensible product decision, but it does mean the vendor's model economics, and any future model change, are entirely theirs to control.
Leading platform for model flexibility & routing in each use case
Picked mechanically: the highest total coverage vendor in each lane that documents full evidence on this axis, one vendor per row. Scores are out of 14.
-
1. Appian, for enterprise operations agents
14 / 14Full enterprise operations agents ranking · Compare the whole lane on all 14 axes
-
2. FLOWX.AI, for multi-agent platforms
14 / 14Full multi-agent platforms ranking · Compare the whole lane on all 14 axes
-
3. Gumloop, for GTM and revenue agents
14 / 14Full GTM and revenue agents ranking · Compare the whole lane on all 14 axes
-
4. Mastra, for agent infrastructure platforms
14 / 14Full agent infrastructure platforms ranking · Compare the whole lane on all 14 axes
-
5. ServiceNow, for customer support agents
14 / 14Full customer support agents ranking · Compare the whole lane on all 14 axes
-
6. UiPath, for agent builders
14 / 14Full agent builders ranking · Compare the whole lane on all 14 axes
-
7. GitHub Copilot, for coding agents
13.5 / 14Full coding agents ranking · Compare the whole lane on all 14 axes
-
8. Dataiku, for data analyst agents
13 / 14Full data analyst agents ranking · Compare the whole lane on all 14 axes
-
9. HappyRobot, for voice agents
13 / 14Full voice agents ranking · Compare the whole lane on all 14 axes
-
10. Torq, for security and SOC agents
12.5 / 14Full security and SOC agents ranking · Compare the whole lane on all 14 axes
-
11. Edge Delta, for SRE and DevOps agents
12 / 14Full SRE and DevOps agents ranking · Compare the whole lane on all 14 axes
-
12. SS&C Blue Prism, for browser and computer-use agents
12 / 14Full browser and computer-use agents ranking · Compare the whole lane on all 14 axes
-
13. Innovaccer, for healthcare agents
11.5 / 14Full healthcare agents ranking · Compare the whole lane on all 14 axes
Documented coverage by use case
Share of each lane documenting full coverage on this axis. Vendors that sit in two lanes count in both, the same rule the rankings and matrices use.
| Use case | Full coverage | Share |
|---|---|---|
| coding agents | 45 of 64 | 70% |
| multi-agent platforms | 35 of 52 | 67% |
| agent builders | 74 of 115 | 64% |
| agent infrastructure platforms | 109 of 186 | 59% |
| browser and computer-use agents | 20 of 42 | 48% |
| SRE and DevOps agents | 14 of 37 | 38% |
| data analyst agents | 19 of 59 | 32% |
| customer support agents | 25 of 113 | 22% |
| voice agents | 18 of 83 | 22% |
| enterprise operations agents | 57 of 297 | 19% |
| security and SOC agents | 16 of 85 | 19% |
| GTM and revenue agents | 10 of 138 | 7% |
| healthcare agents | 2 of 69 | 3% |
Recent verified changes from the vendors named above
Capability coverage is not a static picture. These are the most recent sourced change log entries for the platforms listed above, newest first, one per vendor. Scores on this page update as entries like these are verified.
-
SS&C Blue Prism agent capability
Low impactEvery Blue Prism Assistant agent in WorkHQ now understands requests and replies in the user's own language, including Japanese, Portuguese and Spanish. Until this release only the Product Help and Workflow Builder agents were multilingual. The Assistant interface itself is also now available in Japanese.
September 30, 2026 · Partially Verified · All SS&C Blue Prism changes
-
Gumloop agent capability
Medium impactUsers can now text Gumball, Gumloop's personal agent, over iMessage, RCS or SMS to start a task, get replies in the same thread and approve steps by text. It is a public beta that admins can switch off by role.
September 29, 2026 · Partially Verified · All Gumloop changes
-
FLOWX.AI agent capability
High impactFlowX.AI 5.13.0 adds an AI agent to its Designer that surveys an existing app, drafts a plan from a plain language request, then builds or edits processes, screens, workflows and data types. It runs only after the builder confirms the plan and reviews each step, works on its own branch, and refuses destructive changes such as deleting a process with live instances.
September 28, 2026 · Verified · All FLOWX.AI changes
-
GitHub Copilot agent capability
Medium impactGitHub Copilot code review gained auto-resolution of review comments and updated analysis behavior, so the reviewing agent can close out its own findings rather than only reporting them.
September 11, 2026 · Verified · All GitHub Copilot changes
-
Mastra agent capability
Medium impactMastra introduced Filesystem Skills for Workspaces, enabling reusable skills for multi-agent systems.
September 11, 2026 · Verified · All Mastra changes
Full change log · updated weekly across the whole index
Questions buyers ask
Does a low score mean the platform is locked to one model?
It means the vendor does not document model choice as a buyer facing option. Many such platforms do swap models internally. The point is that you would not know, and you could not choose.
Why do bring your own model and bring your own key matter?
It moves inference cost and data handling onto your own provider account, which changes both the economics and the compliance conversation. It also makes the vendor's markup visible.
Is routing the same as multi model support?
Routing is stronger. Multi model support means you can pick one. Routing means the platform sends each task to an appropriate model, which is where the cost savings usually come from.
The other 13 axes
No single axis decides a shortlist. Buyers who care about this one usually check triggers & channel coverage and apis, sdks & mcp extensibility next, or open the full taxonomy to see how the 14 axes fit together.