# Portkey vs Helicone: which do AI models recommend for LLM gateways, October 2026

IT AI Recommendation Index, October 2026 Edition, LLM gateways and observability. Seven of fourteen models named Portkey first on the direct prompt; one named Helicone. Page: https://it-ai-index.com/it-data/llm-gateways-and-observability/portkey-vs-helicone/

| | First-choice share | Rank | Negative rate | Labels | Models naming it |
|---|---|---|---|---|---|
| Portkey | 31% | #2 of 8 | 11% | 53 | 14 of 14 |
| Helicone | 6% | #4 of 8 | 15% | 27 | 12 of 14 |

## The direct prompt, model by model

- Claude Haiku 4.5: portkey first (first choices: Portkey) (alternatives: Bifrost, LiteLLM, Nexos.ai, OpenRouter)
- Perplexity Sonar: portkey first (first choices: Portkey) (alternatives: Bifrost, LiteLLM)
- DeepSeek V4 Flash: portkey first (first choices: Portkey) (alternatives: Helicone, LiteLLM, OpenRouter)
- Qwen 3.7 Flash: portkey first (first choices: Portkey) (alternatives: Amazon Bedrock, Azure AI, LiteLLM)
- Kimi K2: portkey first (first choices: Portkey) (alternatives: Cloudflare AI Gateway, Kong AI Gateway, LiteLLM)
- GLM 4.7 FlashX: portkey first (first choices: OpenRouter, Portkey) (alternatives: Kong AI Gateway, LiteLLM)
- Muse Glimmer 30B: portkey first (first choices: Portkey) (alternatives: LiteLLM, OpenRouter)
- MiniMax M2.5: helicone first (first choices: Helicone, Portkey.ai) (alternatives: Cloudflare AI Gateway, Kong AI Gateway)
- GPT-5.4 mini: neither first, one named (first choices: LiteLLM) (alternatives: Portkey)
- Gemini 3.5 Flash: neither first, one named (first choices: LiteLLM) (alternatives: Bifrost, Cloudflare AI Gateway, Portkey)
- Grok 4.20: neither named (first choices: Portkey AI Gateway) (alternatives: Bifrost, Cloudflare AI Gateway, LiteLLM)
- Mistral Small: neither named (first choices: Nexos.ai) (alternatives: Cloudflare AI Gateway)
- Llama 4 Maverick: neither named
- GPT-6 Luna: neither named (first choices: Portkey, now branded PRISMA AIRS AI Gateway) (alternatives: Cloudflare AI Gateway, LiteLLM)

## What the models said about Portkey

- "avoid Portkey hosted due to CVEs and black-box concerns" (DeepSeek V4 Flash, negative prompt, hard negative)
- "its paid plans start at $49/month, so it is usually less attractive than LiteLLM or OpenRouter for the tightest budgets" (Perplexity Sonar, budget prompt, soft negative)
- "Acquired by Palo Alto Networks in May 2026, which introduces some roadmap uncertainty" (MiniMax M2.5, negative prompt, soft negative)
- "Go with Portkey if you want an all-in-one platform that's enterprise-ready with strong compliance certifications." (MiniMax M2.5, paraphrase prompt, first choice)
- "If you prefer a managed service so your team can focus strictly on building product features, choose Portkey." (Gemini 3.5 Flash, paraphrase prompt, first choice)
- "Great all-rounder. Excellent balance between enterprise governance, prompt management, and observability." (Gemini 3.5 Flash, scale prompt, first choice)

## What the models said about Helicone

- "What I'd Avoid... Helicone | In maintenance mode since Mintlify acquisition" (Kimi K2, paraphrase prompt, hard negative)
- "now in maintenance mode. Avoid for new builds." (DeepSeek V4 Flash, paraphrase prompt, hard negative)
- "now maintenance mode post-2026 acquisition—no new signups, but OSS viable ... Hosted future uncertain; basic evals." (Grok 4.20, paraphrase prompt, soft negative)
- "Choose Helicone if you need to debug why a specific LLM chain is failing, require the absolute lowest latency" (Qwen 3.7 Flash, comparative prompt, first choice)
- "Start with Helicone. It is the most "agile" tool for a mid-sized B2B company." (Qwen 3.7 Flash, paraphrase prompt, first choice)
- "I'd recommend starting with Portkey.ai or Helicone" (MiniMax M2.5, direct prompt, first choice)

Share is the count of first choices across the direct, paraphrase, budget and scale prompts over all fourteen models, for a mid-market B2B company; rank is within the category. Comparisons are drawn for the top eight products in each category. Published under CC BY 4.0; the output is the models' output, and nothing here is a recommendation by the index.