2 categories · RankedPortkey
Portkey is a production stack for generative AI builders, working as a gateway and observability layer for LLM API calls, including routing, logging, caching and rate limiting · portkey.ai ↗
Strongest at small business, contested with LiteLLM. Ranked 3 of 82 for enterprise buyers.
The standing depends on who is buying. Every model is asked the same questions for each company size.
Ranked first, contested: LiteLLM has 33%.
LiteLLM is first at 36%. 10 of 14 models made it the first choice when asked directly.
Bifrost is first at 23%.
Why models recommend it
“the "best" LLM gateway typically comes down to a choice between the open-source standard, LiteLLM, or the observability-first solution, Portkey”Qwen 3.7 Flash · Direct question · first choice
“consider starting with self-hosted open-source options (like Portkey's MIT-licensed tier or Envoy AI Gateway) where you control the data plane”Kimi K2 · Negative question · first choice
“Portkey has long been the mid-market pick for teams that want guardrails, caching and logging in one product”Claude Haiku 4.5 · Direct question · first choice
“Portkey is my top recommendation for a mid-sized B2B company needing both LLM observability and routing.”Grok 4.20 · Paraphrase question · first choice
What gives them pause
“Avoid percentage-fee gateways until your monthly model spend exceeds $10,000–$35,000, as the fees become cost-prohibitive”MiniMax M2.5 · Budget-constrained question · warning
“Providers like standard versions of OpenRouter or Portkey typically process logs across public infrastructure.”Qwen 3.7 Flash · Negative question · warning
“It eliminates the subscription cost associated with commercial gateways (like Merge.dev or Portkey).”Qwen 3.7 Flash · Budget-constrained question · warning
“Exercise caution with OpenRouter, Portkey, Vercel AI Gateway, and similar services”Perplexity Sonar · Negative question · warning
Also on the shortlist
The products named in the same answers as Portkey, over the 169 answers that named it, and how often each was the first choice instead.
| Product | Named together | Picked instead | |
|---|---|---|---|
| 144 | 36 | Compare → | |
| 87 | 7 | Compare → | |
| 76 | 2 | Compare → | |
| 74 | 24 | Compare → | |
| 69 | 2 | View → | |
| 46 | 3 | Compare → | |
| 36 | 2 | Compare → | |
| 26 | 2 | View → | |
| 21 | 2 | View → | |
| 19 | 1 | View → |
What the models named instead of Portkey →
A side-by-side page exists where both products are among a category's top eight.
At a glance
Stated on Portkey's own site, not checked by Gane.
- For
- AI teams
- Price stated
- Free Forever; $49/month; Custom Pricing
- Starting price
- $49/month
- Free plan or trial
- Get started for free
- Certifications
- HIPAA, SOC 2 Type II, GDPR, ISO 27001
- Customers named
- Figg, Haptik, Qoala, Ario, Internet2, RVO Health, Snorkel AI, QA.tech
How this is measured
Fourteen AI models from 12 labs are asked the same buying questions six ways, for three company sizes. Standings refresh every edition.
Gane sells AI visibility software. No company can pay to change a result.
Read the method →Sources returned with these answers
Fourteen of the fourteen models returned a source list in this edition: five return the pages they cited and nine the search results they consulted. These counts come from 156 of the 169 answers that named Portkey and are not a share of its labels.
Domains cited
387 of the 458 domain citations in answers naming Portkey came from a third party's page. Thirty-four were to the publisher's own pages (the Gane Index), which the models cite like any other source; they are not counted as a third party.
Pages cited
Pages are listed as the models cited them.
By model
| Model | First choice | Alternative | Mention | Warning | Labels |
|---|---|---|---|---|---|
| Claude Haiku 4.5 | 1 | 1 | 0 | 0 | 2 |
| GPT-5.4 mini | 0 | 2 | 0 | 1 | 3 |
| Gemini 3.5 Flash | 3 | 2 | 1 | 0 | 6 |
| Perplexity Sonar | 2 | 2 | 0 | 1 | 5 |
| Grok 4.20 | 3 | 1 | 1 | 0 | 5 |
| Mistral Small | 2 | 0 | 1 | 0 | 3 |
| DeepSeek V4 Flash | 1 | 3 | 2 | 0 | 6 |
| Llama 4 Maverick | 0 | 1 | 1 | 0 | 2 |
| Qwen 3.7 Flash | 2 | 0 | 0 | 2 | 4 |
| Kimi K2 | 2 | 2 | 1 | 0 | 5 |
| GLM 4.7 FlashX | 0 | 3 | 1 | 0 | 4 |
| MiniMax M2.5 | 0 | 1 | 1 | 1 | 3 |
| GPT-6 Luna | 2 | 1 | 0 | 0 | 3 |
| Muse Glimmer 30B | 1 | 2 | 1 | 0 | 4 |
What it publishes
| Kind | Pages | Last 90 days | 2025-10 to 2026-09 | Latest | Categories named |
|---|---|---|---|---|---|
| Report or ebook | 70 | undated | |||
| Blog | 45 | 2 | 2026-09-15 | LLM gateways | |
| Case study | 8 | 8 | 2026-09-15 | ||
| Glossary or explainer | 8 | 0 | 2026-05-04 | ||
| Comparison | 5 | 4 | 2026-09-15 | ||
| Conference or event | 1 | undated | |||
| Template or tool | 1 | 0 | 2023-07-01 |
Most recent
- 2026-09-15Portkey vs litellm comparison
- 2026-09-15All case study
- 2026-09-15Leading delivery platform case study
- 2026-09-15Portkey ai info page blog
- 2026-09-15Openrouter alternatives comparison
- 2026-09-15Litellm alternatives comparison
How it is countedHide how it is counted
Every page portkey.ai exposes, subdomains included. Kind is read from the address and title. The last 90 days, the latest date and the twelve months count pages by when they were published, from the site's feeds, a date in the address, or the page's own publication date, read from up to a hundred of its most recently changed pages; a page that says only when it last changed is counted in its kind but not in when, so the recent counts are a floor, and a kind none of whose pages gives a publication date reads undated. An event counts as online when its address or title says so (webinar, on demand, virtual or online summit); a conference, summit, trade show, expo or roadshow that does not say so is counted as a conference or event, which on a vendor's site is mostly in person. Read September 29, 2026.