Model choice: a census of 143 listings

Model selection happens inside whichever CLI you launch, and the bridge fixes no boss role between them, but the set of bridgeable agents is restricted today to Claude Code and Codex.

Source retrieved 2026-10-03

Work is organised around several vendor stacks rather than one model, and platform choice is advised against the client's existing technology, but no explicit free model choice is promised.

Source retrieved 2026-10-03

You pick the model: one Adapter contract covers hosted, local and custom providers, and router, ensemble and fallback wrappers compose several candidates.

Source retrieved 2026-09-09

You choose the provider and can switch without rewriting agents; 30+ providers behind one API.

Source retrieved 2026-09-02

The product is positioned as a layer for any agent, so the model and the agent are the buyer's choice entirely; Aident sells capabilities, never inference.

Source retrieved 2026-10-03

You choose the model per session with --model and can switch mid-chat with /model, and the vendor claims connectivity to almost any LLM rather than a fixed engine.

Source retrieved 2026-09-30

The buyer picks the foundation model, Aisera's own or a third party's, and the LLM Gateway can also route automatically between open and closed models for cost and latency.

Source retrieved 2026-09-24

You pick the reasoning model from the vendor's fixed per-plan list and it persists across sessions, but auxiliary models are fixed and there is no third-party provider selection.

Source retrieved 2026-09-10

The buyer chooses which model or API to subscribe to from a catalogue of 600+ products; the vendor does not route you to a model of its choosing.

Source retrieved 2026-10-03

The buyer picks the model by picking the client: Argorant is the tool layer behind one MCP endpoint and takes whatever agent you already pay for.

Source retrieved 2026-09-26

Model choice sits with Arsum's architecture work and is expected to change over time; two different providers appear in delivered work, so the client is not locked to one vendor.

Source retrieved 2026-10-03

You choose the LLM per agent from pre-deployed models or add your own provider, and you can equally bring an entirely external conversational AI platform instead.

Source retrieved 2026-09-24

You pick the model per expert, session or message; Prism routing is an opt-in alternative.

Source retrieved 2026-09-02

You choose the model per agent, and Microsoft states teams can mix heterogeneous models to match capability or cost for different roles.

Source retrieved 2026-09-19

Model choice is claimed for the chatbot line through an in-house routing gateway, and the automation page says they build on the AI models you already use; the agent page states no model policy.

Source retrieved 2026-09-11

You choose the models and can mix them across one workflow, per bot or per Autonomous Node; Botpress then routes between your own fast, best and fallback slots, and picks for you only if you set nothing.

Source retrieved 2026-09-02

Choice arrives through bring-your-own-key rather than a published picker: the agency page claims support for all LLMs with your own API key, but no selectable model list or default model is documented.

Source retrieved 2026-09-13

You pick the model in Stagehand config, or omit it and let the gateway choose per call.

Source retrieved 2026-09-02

You pick the agent's LLM and can swap it in the Playground or API; Cartesia owns the speech models and manages the LLM integration.

Source retrieved 2026-09-02

You choose the model per session and per subagent, but only from Claude models.

Source retrieved 2026-09-02

You choose the model and effort per session or per subagent, but only within Claude; admins can pin the allowlist.

Source retrieved 2026-09-02

You choose the model and the provider, and the vendor frames that as its differentiator; the same client talks to more than thirty providers plus local runtimes.

Source retrieved 2026-09-12

Hosted plans give no model picker, but Enterprise self-hosted deployments may point CodeRabbit at their own LLM provider or account for inference.

Source retrieved 2026-09-15

You pick the model per run and can define custom providers, but the first-party catalogue is OpenAI models only.

Source retrieved 2026-09-02

Inference is routed through OpenRouter and the agent record stores a selected model, but the vendor publishes only a basic/premium split rather than a list you choose from.

Source retrieved 2026-09-28

You choose the model per agent, per task or per crew; five providers have native SDKs and everything else routes through LiteLLM or any OpenAI-compatible base URL.

Source retrieved 2026-09-13

You pick the model per request, or hand selection to Auto/Cursor Router; each choice bills at that model's own rate.

Source retrieved 2026-09-02

You choose the STT model, the LLM provider and model, and the TTS voice, and can swap the LLM or voice mid-session with UpdateThink and UpdateSpeak messages.

Source retrieved 2026-09-02

You pick the model in Devin CLI, or leave it to the Adaptive router; cloud Devin sessions document no model selector.

Source retrieved 2026-09-02

You pick the model per conversation, or hand it to the Adaptive router; Enterprise admins can restrict the selectable set by model or by provider.

Source retrieved 2026-09-03

You pick the model per app and per node, from bundled providers or any provider you install and key yourself; workspace defaults are configurable.

Source retrieved 2026-09-02

The buyer picks the model per agent and nothing is gated behind a higher plan, though admins may cap which model tiers members can select and costlier models burn more credits.

Source retrieved 2026-10-01

You pick the model per agent, plus a backup cascade; LLM usage is billed on top of the plan and varies by model.

Source retrieved 2026-09-02

You choose between three vendor-defined models at three credit prices, so there is real choice, but the catalogue is entirely Faceless's own with no third-party model option.

Source retrieved 2026-10-03

You pick the model per session, per subagent and per exec run with -m/--model, or add custom endpoints.

Source retrieved 2026-09-02

You choose the inference provider only on the self-hosted stack; the agent endpoint exposes a model parameter, but Cloud scraping otherwise uses Firecrawl's managed provider path.

Source retrieved 2026-09-24

The vendor picks the model per task and users cannot choose one per chat; the only customer control is a Teams or Enterprise admin switching whole providers off, and at least one provider must stay on.

Source retrieved 2026-09-14

Only a named subset of products exposes a by-name model picker; everywhere else you choose a Standard or Ultra quality tier and Genspark decides which model actually runs.

Source retrieved 2026-09-15

You pick from a vendor-curated catalog, and the plan cards list model selection as a Pro addition; bring-your-own-key is the only route to a model GitHub does not host.

Source retrieved 2026-09-02

Admins choose which of 40+ models are available and to whom; auto routing can pick per task instead.

Source retrieved 2026-09-02

You choose the model per agent, including a router that picks between several at runtime; Gemini is the default path but not a requirement of the framework.

Source retrieved 2026-09-15

You choose the provider and the model, can switch either mid-session with /model, and can register custom OpenAI-, Anthropic- or Ollama-compatible providers from a JSON file.

Source retrieved 2026-09-16

You choose the provider and model per role through FAST_LLM, SMART_LLM and STRATEGIC_LLM config keys rather than accepting a vendor default.

Source retrieved 2026-09-16

The buyer picks the model per agent from 35+ options on both plans, can swap freely, and Enterprise admins can restrict the allowed set.

Source retrieved 2026-09-30

You choose a performance tier from a stack HeyRik manages and reads back through the API, rather than picking a named model; the vendor operates the conversational intelligence itself.

Source retrieved 2026-09-10

Guard sits in front of whichever harness and model you already use, applying one cross-agent policy rather than choosing a model.

Source retrieved 2026-09-29

The vendor calls itself model-agnostic but picks the defaults from its own leaderboard and routes on the buyer's behalf, so model selection is a vendor-configured choice rather than a self-serve setting.

Source retrieved 2026-09-24

You choose the provider and the model, and can switch model and thinking level mid-workflow; Juggler sells no inference of its own, so the choice is yours in every case.

Source retrieved 2026-09-17

The buyer picks the model per conversation from a dropdown spanning the vendor's own models and third-party ones, rather than being routed silently by Julius.

Source retrieved 2026-09-17

You choose, not the vendor: four provider classes are supported and a flag forces a specific one, though leaving the default lets JetBrains pick on price-quality grounds.

Source retrieved 2026-09-18

You pick the model; Auto Model routing is optional and never silently substitutes.

Source retrieved 2026-09-02

You choose the provider everywhere except Fleet's Fast, Pro and Max tiers, where LangChain selects and maintains the underlying model for you; custom configurations can override even those.

Source retrieved 2026-09-18

You choose the provider and model per component, and the documentation states plainly that the framework does not tie you to particular LLMs or vector stores.

Source retrieved 2026-09-25

You choose the model. LangGraph does no model abstraction of its own and the vendor markets provider-independence; the featured integration table spans OpenAI, Anthropic, Google, Bedrock, Ollama and more.

Source retrieved 2026-09-02

Inside the app the vendor picks the models and bills them against an AI budget; over MCP you supply the model, because classification and drafting happen in your assistant and Lead Scorer only stores the result.

Source retrieved 2026-09-18

The buyer picks the underlying providers rather than accepting a fixed stack, with the platform orchestrating them into one agent; note the security page separately claims every component including model deployment is managed in-house.

Source retrieved 2026-09-09

You pick the model and can switch mid-conversation; memory travels with the agent across providers.

Source retrieved 2026-09-02

The buyer picks the model per task from a dropdown and can set a default, and Lindy also picks one itself to control credit spend; no tier restricts which models are reachable.

Source retrieved 2026-09-19

You pick every stage. STT, LLM and TTS are swappable through the plugin system, and realtime speech-to-speech models can replace the three-stage pipeline.

Source retrieved 2026-09-02

v1 API lets you name the model per page; the recommended v2 tier system picks the model for you.

Source retrieved 2026-09-02

Provider choice is an explicit design property: agents, flows and connectors are meant to survive a swap between providers or any compatible API.

Source retrieved 2026-09-09

The platform is deliberately model-agnostic but the vendor does the choosing; users only pick between Manus's own tiers, and the Lite tier is all a free account gets.

Source retrieved 2026-09-19

You choose the model per agent, per request or as a fallback chain; the router is explicitly positioned as provider-agnostic.

Source retrieved 2026-09-02

Maven claims its own architecture is LLM-portable on the platform page, but the contract shows the vendor picks the models and may change them on notice; no buyer-facing model selector is published anywhere.

Source retrieved 2026-09-03

You choose the model: eleven providers are documented and, in .NET, any inference service implementing the standard chat-client interface can back an agent without framework changes.

Source retrieved 2026-09-19

You choose the model per block from a large multi-publisher catalogue, can compare models side by side, and can plug in your own keys or endpoints instead.

Source retrieved 2026-09-20

The built-in AI Agent gives no model picker, but you can route conversations to a Dify or Coze bot you built yourself, including a self-hosted Dify deployment, where the model choice is yours.

Source retrieved 2026-10-03

Builders may name a model per LLM Action from a fixed list, but the assistant's core routing is Moveworks' own gateway choosing the model, and org provider settings can narrow the list.

Source retrieved 2026-09-20

You pick among Muse Spark versions per run or mid-session, and the reasoning docs imply other providers exist, but no alternative provider is named or configured anywhere in the Muse Code docs.

Source retrieved 2026-09-20

You pick the provider and model per node or per agent, and swap it without rebuilding the workflow; a fallback model can be attached to an agent.

Source retrieved 2026-09-02

Model independence is advertised as an architectural principle, set against what they call the fragility of building on a single vendor's API.

Source retrieved 2026-09-26

You choose the speech-to-text, LLM and text-to-speech provider per agent and may mix them freely; the vendor does not pin you to one stack.

Source retrieved 2026-09-09

You choose the model per agent, per run or globally, and any OpenAI-compatible endpoint can be substituted by setting a base URL and key; adapters extend this to further providers.

Source retrieved 2026-09-20

You pick the model: provider, authentication and execution runtime are separate layers, and per-agent or per-model runtime policy decides which harness executes a turn.

Source retrieved 2026-09-21

You choose the provider and model, per agent if you want, and the vendor's own gateway is stated to be optional.

Source retrieved 2026-09-21

You choose the model on every tier; the vendor lists model agnosticism as a feature of the free local plan itself.

Source retrieved 2026-09-21

Paid consumer plans set a preferred model, and the API takes an explicit model id across multiple provider catalogues; otherwise Perplexity picks per query.

Source retrieved 2026-09-02

You choose the model per agent and can swap it at any time without rebuilding the agent, and the vendor makes not being locked to one provider an explicit selling point.

Source retrieved 2026-09-09

The Engine, not the buyer, chooses a model per task; the vendor's claim is portability of the task definition across models rather than user-level model selection.

Source retrieved 2026-10-03

You choose the model per channel in voice and chat configuration; Raven is the recommended default rather than the only option.

Source retrieved 2026-09-02

You choose the model, not the vendor: an agent is portable across LLM vendors by swapping the model string, and the docs state no flagship feature is locked to one vendor.

Source retrieved 2026-09-21

You pick the model per agent, set a fallback model from another provider, and the feature table marks "Bring your own LLM" available on every published tier.

Source retrieved 2026-09-22

Manual model choice exists only in the paid modes on paid plans; Free Mode and Starter always defer to Replit's router.

Source retrieved 2026-09-22

You pick the conversation model per agent from a published, per-minute-priced list, and you can substitute your own model entirely through the custom-LLM WebSocket.

Source retrieved 2026-09-23

Rox meters by model class and lists six AI/inference providers as subprocessors, so it is multi-model internally, but the buyer never chooses the model; own-model integration is a bespoke conversation.

Source retrieved 2026-09-03

The vendor advertises freedom from provider lock-in across three commercial models, but the model is settled inside a written scope the agency controls rather than switched by the buyer.

Source retrieved 2026-09-28

The customer selects the model option in Setup and can override it per agent or subagent in Agent Script; leaving the default hands the choice to Salesforce's managed mix.

Source retrieved 2026-09-03

The model lives in your own client, which you bring and pay for separately, so the buyer picks the model rather than the vendor.

Source retrieved 2026-09-25

The vendor picks the default stack; your choice is limited to flagging a workspace EU-sensitive or supplying your own Anthropic/OpenAI key, not free model selection.

Source retrieved 2026-10-03

You pick the model per Agent block across 15+ providers; the default is only a default.

Source retrieved 2026-09-03

Model choice is entirely the buyer's: flow carries audio and text only, and states plainly that you keep your choice of LLM.

Source retrieved 2026-09-26

Bots are instruction sets you run in whichever assistant you already pay for, so the model is your choice there; the model behind the hosted web audit is never disclosed.

Source retrieved 2026-10-03

The buyer picks provider and model per agent node and can switch later from a dropdown; admins may narrow that list but the vendor does not fix one model.

Source retrieved 2026-09-09

You pick the model per agent, but only from Synthflow's list of GPT and Synthflow LLM options.

Source retrieved 2026-09-03

The buyer picks the provider credential and pins a model per agent; the vendor ships no model of its own and admins restrict access per person or team.

Source retrieved 2026-10-03

The extension lets the user pick provider and model, but for agency-built agents and receptionists model selection is TecAdRise's implementation choice and is nowhere published.

Source retrieved 2026-09-12

You choose the telephony carrier, the STT engine, the TTS voice and the execution pipeline, but every LLM option in the calculator comes from one vendor, so model choice stops at OpenAI.

Source retrieved 2026-09-16

You choose the model per assistant and can point it at any public OpenAI-compatible endpoint such as FastRouter, Azure OpenAI, Bedrock, Baseten or vLLM, instead of the vendor picking for you.

Source retrieved 2026-09-09

You pick the model per agent from the Properties panel, models are provisioned through the AI Trust Layer, and the vendor states a model can be swapped later without rebuilding the agent.

Source retrieved 2026-09-03

You choose transcriber, LLM and voice independently and can swap any of them.

Source retrieved 2026-09-03

The buyer picks the model: each Agent carries its own chosen AI model as a setting, drawn from a catalog spanning Anthropic, OpenAI, Google, DeepSeek, Qwen and Mistral.

Source retrieved 2026-09-27

You pick the model per workspace, per conversation and per action type through inference profiles and action overrides, with custom profiles for providers not built in.

Source retrieved 2026-09-27

The buyer picks the STT, LLM and TTS provider and model independently for each agent; the cheapest paid plan includes all of them.

Source retrieved 2026-09-03

You pick from the model picker or write your own routing rules; an admin or your plan can grey out models, and Auto hands selection back to Warp if you want it.

Source retrieved 2026-09-03

You pick the provider on the command line from a published list, and two of them inherit whatever model your browser profile has selected.

Source retrieved 2026-09-21

Any MCP-capable client and model can call it, and the vendor lists several competing agents rather than binding you to one vendor's model.

Source retrieved 2026-10-03

You choose the model from a selector and can route through Zed hosting, your own API keys, an existing subscription, a gateway or a local server.

Source retrieved 2026-09-29

You pick the model per chat and per workflow step from a published catalogue, and can switch mid-task; Auto routing is one option rather than the only path.

Source retrieved 2026-09-09

You choose the provider and model per run or per swarm agent, and you can register entirely new providers and endpoints yourself in a models file.

Source retrieved 2026-09-28

Model selection is presented as an engineering decision made against the client's accuracy, latency and cost constraints rather than a fixed vendor allegiance, including multi-model routing.

Source retrieved 2026-09-25