ai-toolkit/concentrate-ai

Value Proposition & Features

Concentrate AI is a fee-less enterprise AI gateway that provides “one API for every LLM,” giving teams centralized routing, governance, and observability across multiple model providers without adding a gateway markup on token spend or card processing fees. [^7iabid] [^2se8xd] [^v7mbtn] [^dpm62i] [^ow9tin] [^jlfu8e] [^bwves5] [^9jnnhj] It positions itself as a privacy‑first orchestration platform for routing, monitoring, and optimizing AI inference and API traffic, targeting production workloads across industries. [^8vd1in] ^[9ro4ah] The core value proposition is that teams can access leading models (e.g., GPT 5.6, Claude 5 variants, Gemini 3.1 Pro, Grok 4.20) through a normalized API while getting enterprise controls and cost tracking “for free” on top of provider pricing. [^7iabid] ^[9ro4ah] [^dlf1m7] [^dpm62i] [^ow9tin] [^0g24i5]

Core feature descriptions (2–3 sentences each)

Unified multi‑model API. The Concentrate AI Responses API exposes a single base URL (https://api.concentrate.ai/v1) and a normalized interface to interact with “multiple AI model providers,” including GPT 5.6, Claude Opus 5, Claude Fable 5, Gemini 3.1 Pro, Grok 4.20, and others through one API. [^7iabid] This unification means application code points at the gateway instead of each provider’s SDK, with rate limits enforced per API key and subscription tier. [^7iabid] [^77dde6]
Provider routing and automatic failover. Concentrate gives teams “provider routing and automatic fallbacks,” so requests can be sent across providers based on rules and transparently retried when a provider degrades. [^2se8xd] External commentary describes AI gateways similarly as a routing and failover layer that abstracts providers and moves traffic between models without code changes, aligning with Concentrate’s positioning as an LLM gateway for AI teams. [^66jyqi] [^6zxtwd] [^dla05r] [^79dfts]
Spend controls, logging, and governance. The service offers “logs, spend controls, audit trails, data redaction, and access controls,” allowing teams to monitor usage and enforce budgets and policy across models and users from one place. [^2se8xd] [^8vd1in] [^6zxtwd] [^xse25r] [^79dfts] Its privacy policy emphasizes routing, monitoring, and optimizing AI inference, and references SOC 2 Type II and a privacy office, reinforcing a governance‑oriented design. [^8vd1in]
Bring Your Own Key (BYOK). The BYOK feature “lets you store your own provider API keys in Concentrate,” using the gateway as a control layer while keeping spend and accounts under the customer’s ownership. [^h7p0kj] Linked posts stress that Concentrate charges no platform, BYOK, or service fees on token spend, suggesting BYOK is priced at $0 gateway markup. [^dpm62i] [^jlfu8e] [^bwves5] [^9jnnhj] [^0g24i5]
No‑fee pricing model. Multiple launch posts state Concentrate is “the first no‑fee AI gateway” and “the first fee‑less AI gateway,” having “removed all platform fees and credit‑card fees” so customers “pay for the model usage itself, nothing on top.” [^v7mbtn] [^dpm62i] [^ow9tin] [^jlfu8e] [^bwves5] [^eum9g0] Commentary from a third‑party finance outlet notes that Concentrate “slash[ed] pricing directly to the cost price of the AI models themselves, even eliminating credit card processing fees.” [^htvwy6]
Enterprise‑grade orchestration and multi‑tenant support. Launch materials emphasize “enterprise gateway controls” and “everything required to run AI in production, hosted and ready to use,” including unlimited users and unlimited teams, indicating multi‑tenant, enterprise workloads such as financial services, retail, telecom, and healthcare. ^[9ro4ah] [^2se8xd] [^dlf1m7] The privacy policy frames Concentrate as a “privacy‑first orchestration platform” with SOC 2 and DPO support for regulated environments. [^8vd1in]
High‑volume routing performance. A recent benchmark claim states Concentrate processed “100Bn tokens in 8 hours” and “just processed more Gemini 3.5 Flash tokens in 12 hours than OpenRouter has done in the last 5 days combined,” indicating focus on large‑scale routing performance. ^[0n9h7x] [^wo1nbj] This sits within a broader market narrative where model routers and gateways manage tens or hundreds of billions of tokens for high‑volume production AI. ^[p5ih0b] ^[xmj27g] ^[wkmsm7] ^[jcmjs2] ^[7cwr3h]

Feature list (5–8, priority order)

One unified API for many leading LLMs (GPT 5.6, Claude 5 variants, Gemini 3.1 Pro, Grok 4.20, etc.). [^7iabid]
Provider routing and automatic fallbacks for reliability across multiple providers. [^2se8xd] [^66jyqi] [^6zxtwd] [^dla05r]
Fee‑less gateway model with 0% platform, BYOK, service, and card fees on token spend. [^v7mbtn] [^dpm62i] [^ow9tin] [^jlfu8e] [^bwves5] [^9jnnhj] [^htvwy6] [^0g24i5] [^eum9g0]
Spend controls, logs, audit trails, and access controls for enterprise governance. [^2se8xd] [^8vd1in] [^6zxtwd] [^xse25r] [^79dfts]
Bring Your Own Key (BYOK) support, storing customer provider keys in the gateway. [^h7p0kj] [^dpm62i] [^bwves5]
Privacy‑first orchestration with SOC 2 Type II available under NDA and a named DPO. [^8vd1in]
High‑volume routing scale, including processing on the order of 100 billion tokens within hours. ^[0n9h7x] [^wo1nbj] ^[p5ih0b] ^[7cwr3h]
Unlimited users and teams for multi‑team, multi‑environment enterprise use. [^2se8xd] ^[9ro4ah]

Screenshots

No reliable source found. Public, clearly identified official product screenshots for Concentrate AI’s UI were not found in the material reviewed. ^[d8fzly] ^[x3rry3] [^y7fpa3] [^wo1nbj]

Product Roadmap / Announcements

As of August 23, 2026,

2026‑08‑10 – Launch of Concentrate AI as a no‑fee enterprise AI gateway. “We’re launching Concentrate AI to give every team the ability to switch between models with enterprise gateway controls, for free… Concentrate AI is the first no-fee AI gateway,” with features including any model through one API, routing and automatic fallbacks, logs, spend controls, audit trails, data redaction, access controls, unlimited users/teams, and 0% platform and card fees. [^dc8yk4] ^[9ro4ah] [^2se8xd]
2026‑08‑10 – Elimination of platform, BYOK, service, and credit‑card fees. Posts by leadership state Concentrate “has removed all platform fees and credit-card fees,” is “the first fee-less AI gateway,” and customers “pay for the model usage itself” with “no platform fee on token usage and no card-processing fee passed on to customers.” [^u6lnc2] [^eum9g0] [^v7mbtn] [^dpm62i] [^jlfu8e] [^bwves5] [^9jnnhj]
2026‑08‑10 – BYOK emphasis and cost‑price routing statement. A finance summary reports that, facing competition, Concentrate “slash[ed] pricing directly to the cost price of the AI models themselves, even eliminating credit card processing fees,” reinforcing BYOK and fee‑less routing as a central product stance. [^htvwy6]
2026‑08‑16 – Documentation update for Responses API. The API Introduction was last updated mid‑August 2026, describing the Responses API, supported models, and subscription‑tier rate limits, indicating ongoing platform and documentation iteration. [^7iabid] [^77dde6]

Recent Developments (past 90 days)

2026‑08‑19 – Market validation post about Stripe’s acquisition of OpenRouter. CEO Ari Jacoby commented that Stripe’s acquisition of OpenRouter for $8B “validates [the] AI router market” and noted Concentrate AI as an independently owned at‑scale AI gateway focused solely on being the best gateway for companies. [^b1vzm0] ^[jm9ql8]
2026‑08‑16–20 – Third‑party commentary on AI gateway competition and pricing. A finance article frames Concentrate AI’s move to charge only provider model cost and eliminate card fees as a response to intensifying AI routing competition and large acquisition valuations. [^htvwy6] ^[xmj27g] ^[jcmjs2]
2026‑08‑10–14 – Public messaging on fee‑less gateway positioning. Multiple LinkedIn posts from leadership repeatedly emphasize that Concentrate has removed all gateway fees and card surcharges, positioning it as “the first fee-less AI gateway” and “the first and only AI Gateway” to remove service and card fees on token spend. [^u6lnc2] [^eum9g0] [^v7mbtn] [^dpm62i] [^jlfu8e] [^bwves5] [^9jnnhj] [^a8cg8k]
2026‑07‑27 – Performance claim on token routing scale. A post highlights that Concentrate AI processed “100Bn tokens in 8 hours,” and another notes it processed more Gemini 3.5 Flash tokens in 12 hours than OpenRouter did in five days, presenting it as a high‑scale router in the “AI token router wars.” ^[0n9h7x] [^wo1nbj] ^[p5ih0b] ^[7cwr3h]

History and Origin Story

Concentrate AI, Inc. is incorporated in Delaware with a registered address at 1201 N Market St, Suite 200, Wilmington, Delaware, and operates as a privacy‑first orchestration platform for routing and optimizing AI inference. [^8vd1in] Public launch materials on July 28 and August 10, 2026 describe Concentrate AI as an LLM gateway for AI teams emerging from stealth, with Ari Jacoby identified as CEO and founder and Todd Lieberman describing the company as a continuation of a two‑decade history of building aggregation businesses that “bring together” providers, now focused on AI inference. [^66jyqi] ^[7vhw6q] ^[uycv4b] ^[9ro4ah] [^u6lnc2] Posts from the founding team mention prior experience in aggregation businesses (VoiceStar, Circulate, Deduce) and frame Concentrate as the next step in that pattern, consolidating AI inference providers behind one gateway API. ^[uycv4b] [^z8cykr] The company’s early phase centers around its fee‑less gateway launch and positioning itself as an independent, at‑scale gateway in a consolidating AI infrastructure market where large competitors like OpenRouter are being acquired. [^u6lnc2] [^eum9g0] [^b1vzm0] ^[jm9ql8] ^[jcmjs2] ^[xmj27g]

Notable Team Members

Ari Jacoby – CEO / Founder. Ari Jacoby is identified in LinkedIn posts as CEO of Concentrate AI, with external commentary noting he “comes out of stealth mode and debuts Concentrate AI” as an LLM gateway for AI teams, and he appears on fundraising panels discussing raising a seed round for Concentrate AI. ^[7vhw6q] [^66jyqi] [^0p8sko] [^b1vzm0] His prior background includes founding multiple companies with four exits, and he is positioned as the primary public face driving the company’s go‑to‑market and fundraising narrative. [^0p8sko] ^[uycv4b]

Todd Lieberman – Co‑founder / founding leadership. Todd Lieberman publicly states “I started another company… now, at Concentrate AI, we are bringing together AI inference providers,” describing a 20‑year history of building aggregation businesses (VoiceStar, Circulate, Deduce) and positioning Concentrate as the latest aggregator in that sequence. ^[uycv4b] [^z8cykr] His commentary emphasizes using purchasing scale to remove platform and card fees from the gateway and presents him as a core part of the founding leadership team focused on strategy and economics. ^[uycv4b] [^z8cykr] [^bwves5]

Zach Moskow – Founding team, Head of GTM & Product. Zach Moskow is described as “Founding Team and Head of GTM & Product at Concentrate AI,” with experience across strategy, operations, AI, and product growth, and plays a key role in launching the “first fee-less AI gateway” and messaging the elimination of platform and card fees. ^[5ygazf] [^u6lnc2] [^v7mbtn] [^9jnnhj] His posts focus on gateway economics, customer cost savings, and event speaking engagements representing Concentrate AI. ^[5ygazf] ^[5rfc7e]

Shannon Gelson – Data Protection Officer (DPO). The privacy policy names Shannon Gelson as Data Protection Officer for Concentrate AI, with responsibility for privacy oversight and a dedicated privacy office address, indicating a leadership role in compliance and data protection. ^[q678l3] [^8vd1in]

Market Sizing

Category, Market Size, and Category Growth

Concentrate AI fits within the enterprise AI gateway / LLM gateway / AI model router category: a control layer that provides one API, routing, failover, observability, spend controls, and governance across multiple LLM providers for production applications. [^dlf1m7] [^66jyqi] [^6zxtwd] [^dla05r] [^xse25r] [^79dfts] Analyst data on the Enterprise AI Gateway Market estimates a market value of USD 0.88B in 2025, projected to reach USD 11.32B by 2035 with a 29.12% CAGR, with LLM gateways dominating the segment at 42.8% share in 2025. ^[d00m9w] A broader trend report suggests AI infrastructure layers (including gateways, observability, evaluation) could represent 5–10% of a $300B AI software market by 2027, or roughly $15–30B annually, underscoring significant growth potential for gateway vendors like Concentrate. ^[jyaju0] ^[brl32h] ^[p5ih0b]

Pricing

Public statements emphasize shape rather than a tier matrix; there is evidence of subscription‑tier rate limits but no detailed tier names or prices.

Tier / Plan	Price (gateway)	Notes
Gateway platform fees	$0	Concentrate “has removed all platform fees” on token spend.
BYOK / service fees	$0	“You pay $0 in platform, BYOK or service fees on your token spend.”
Credit‑card surcharges	$0	“You also pay $0 in credit-card processing surcharges.”
Subscription tiers (rate limits)	Not disclosed	API docs note rate limits “based on your subscription tier,” but tier details and prices are not publicly documented.

Sources for Table: [^dpm62i] [^eum9g0] [^v7mbtn] [^9jnnhj] [^jlfu8e] [^bwves5] [^htvwy6] [^77dde6] [^7iabid]

Beyond the above fee structure, no public pricing table with named plans and per‑month or per‑token gateway fees is available. [^dpm62i] [^ow9tin] [^0g24i5] [^77dde6] [^htvwy6]

Revenue Trajectory Estimates

No reliable source found. There are no public figures for Concentrate AI’s revenue or ARR; available information focuses on pricing structure and market validation rather than financial performance. [^dpm62i] [^ow9tin] [^htvwy6] [^b1vzm0]

Competitive Landscape

Who it’s for, who it’s not for

Concentrate AI is for teams running high‑volume, multi‑provider AI workloads that need an enterprise gateway with routing, failover, observability, spend controls, BYOK, and governance, but want to avoid gateway markups and card surcharges on their token spend. [^2se8xd] [^dlf1m7] [^8vd1in] [^6zxtwd] [^dla05r] [^xse25r] [^79dfts] [^dpm62i] [^ow9tin] [^htvwy6] This includes engineering, data, and product teams in sectors like finance, retail, telecom, and healthcare that already consume multiple LLMs and care about centralized policy, privacy, and compliance while managing large token volumes. ^[9ro4ah] ^[p5ih0b] ^[d00m9w] ^[jyaju0] ^[brl32h]

It is less suited for very small projects or single‑provider prototypes that do not need multi‑provider routing, enterprise governance, or high‑volume cost controls, and for teams whose needs are fully met by provider‑native APIs or tightly integrated platform gateways (e.g., Cloudflare, Vercel) and who prefer those ecosystems. ^[dv46x4] [^vpn54u] [^ophi80] [^79k687] It may also be a weaker fit for organizations that require self‑hosted, open‑source gateways they operate themselves, since Concentrate positions itself as a hosted enterprise gateway rather than a self‑managed OSS proxy. [^dlf1m7] [^dla05r] ^[dv46x4] [^wyrh7u] [^vpn54u]

Viable Alternatives

OpenRouter – Managed AI marketplace and gateway with hundreds of models, unified billing, and auto‑routing, but typically charges a platform fee on credits; positioned as a leading alternative for teams that want one key and are comfortable with per‑token markups. ^[xmj27g] ^[dv46x4] [^vpn54u] [^ophi80] [^an9syd] [^c29grj] ^[jcmjs2] ^[wkmsm7]
TrustedRouter
LiteLLM – Open‑source proxy and LLM gateway that can be self‑hosted, providing unified interfaces, budget tracking, and routing across many providers for teams that want control and zero SaaS gateway markup. ^[dv46x4] [^wyrh7u] [^c8xvzs] [^vpn54u] [^ophi80] [^an9syd] [^c29grj] ^[7cwr3h]
Portkey – Managed + self‑hostable AI gateway focused on routing, observability, budgets, and guardrails, with paid managed plans starting around tens of dollars per month and an OSS tier for self‑hosting. ^[dv46x4] [^wyrh7u] [^vpn54u] [^ophi80] [^gb7t1f]
Cloudflare AI Gateway – Managed gateway integrated with Cloudflare’s edge stack, offering caching, rate limiting, dynamic routing, DLP, and guardrails, suited to teams already on Cloudflare. ^[dv46x4] [^vpn54u] [^ophi80] [^c8xvzs] [^an9syd]
Kong AI Gateway / TrueFoundry / Vercel AI Gateway – Enterprise and developer gateways embedded into broader API or deployment platforms, providing routing, failover, and governance for organizations standardized on those ecosystems. [^79itw6] [^wyrh7u] [^vpn54u] [^ophi80] [^79k687] [^ykbt0e] [^c29grj] [^sfbx1h]

Competitor Table

Competitor	Description
OpenRouter	Managed AI marketplace and gateway offering a unified API and billing for hundreds of models, with auto routing and a per‑credit platform fee, widely cited as a leading LLM gateway.
LiteLLM	Open‑source LLM gateway and proxy that teams self‑host to unify provider APIs, routing, budgets, and observability without SaaS gateway fees.
Portkey	AI gateway with routing, observability, guardrails, and budget controls, available as OSS and managed SaaS, often recommended for production LLM and agent apps.
Cloudflare AI Gateway	Cloudflare‑integrated AI gateway that sits at the edge, providing caching, rate limiting, dynamic routing, and data protection for teams already using Cloudflare services.
Kong AI Gateway	AI gateway built into Kong’s API platform, offering routing, governance, and observability for organizations standardizing their APIs on Kong.
TrueFoundry AI Gateway	Gateway bundled with model deployment and governance features such as virtual models, RBAC, budget limits, and routing strategies, targeting enterprises needing standardized AI operations.
Vercel AI Gateway	Gateway integrated with Vercel’s frontend and serverless tooling, with zero token markup and deep SDK support, designed for LLM apps hosted on Vercel.

Sources for Table: ^[xmj27g] ^[dv46x4] [^vpn54u] [^ophi80] [^an9syd] [^c29grj] ^[jcmjs2] ^[wkmsm7] [^wyrh7u] [^c8xvzs] ^[7cwr3h] [^gb7t1f] [^79itw6] [^79k687] [^ykbt0e] [^sfbx1h]