Internal developer portals
Asked as “internal developer portal”, and as “internal developer platform”, on behalf of a mid-market B2B company. 53 first choices recorded across the direct, paraphrase, budget and scale prompts, twelve models each. Added to the September 2026 Edition; its answers are read by the distilled judge (ai-indexes-judge-qwen3-14b-run3), not the claude-opus-5 judge of the earlier categories: how the two compare.
Standing · first-choice share
53%
Clear leader
53Port23Backstage08OpsLevel17others
53% of first choices, clear leader.
By buyer segment
The same question asked on behalf of a different buyer. Each standing is computed within its segment; they sit side by side and are never added together.
The standing
Share is the count of first choices across the direct, paraphrase, budget and scale prompts, over all twelve models, for a mid-market B2B company. Ordered by share.
ProductFirst-choice shareNegative rateLabelsQuadrant
Show the four products at 0%, ordered by negative rate
Bars are the share of first choices, 0 to 100Every product with at least 10 labels here. Every product name links to its product page.
Recommended versus criticized
Every product with at least 10 labels here, on both axes. The 30% line names a quadrant, not the verdict above: that one needs more than 40%.
Criticized challengerCriticized default
Negative label rate →
01
02
03
04
05
06C07
08
09Accepted challengerEndorsed leader
0%First-choice share → · lines at 30% share and 25% negative70%
Key
01Port53%
Port53%02Backstage23%
Backstage23%03Palo Alto Networks Cortex Cloud8%
Palo Alto Networks Cortex Cloud8%04OpsLevel8%
OpsLevel8%05Humanitec2%
Humanitec2%06Roadie0%
Roadie0%07Configure80%
08Cycloid0%
Cycloid0%09Atlassian Compass0%
Atlassian Compass0%What they warned about
Five of twelve models held their first choice under the paraphrase. GPT-5.4 mini, Perplexity Sonar, Mistral Small, Llama 4 Maverick, Qwen 3.7 Flash, Kimi K2 and GLM 4.7 FlashX changed. A high negative share on a product with few labels is a warning. A low share on a product with many labels is salience, not sentiment.
Backstage42%
28 of 67 labels negative · 11 of 12 models · 4 hard negative
“Strongly Avoid ... Backstage is a framework, not a finished product. Many organizations treat it as such and end up disappointed.” GLM 4.7 FlashX, negative prompt
OpsLevel12%
5 of 41 labels negative · 4 of 12 models
“Most proprietary portals (Port.io, OpsLevel, Atlassian Compass, etc.) carry **vendor lock-in risks**” DeepSeek V4 Flash, negative prompt
Palo Alto Networks Cortex Cloud11%
5 of 47 labels negative · 4 of 12 models
“No public pricing (sales-led), so harder to budget for — generally not ideal if budget transparency matters.” DeepSeek V4 Flash, budget prompt
Port7%
4 of 59 labels negative · 4 of 12 models
“Port, if fully self-hosted/on-prem deployment is a hard requirement... confirm that those meet your actual data-residency and isolation requirements before committing.” GPT-6 Luna, negative prompt
What they cite
Citations exist only for the models that return a source list: fourteen of the twelve in this edition, and all six flagship models on the expanded tier.
Sites the answers cite
78 of 84 answers in this category came back with a source list, from 14 of 12 models: citations where the model returns them, or the search results it consulted. 1041 links across 172 sites, every framing counted. Ranked by the number of answers carrying the site or page.
vendor site · Port47 answers · 83 citations · 13 models
vendor site · Cycloid40 answers · 62 citations · 11 models
34 answers · 34 citations · 12 models
33 answers · 33 citations · 12 models
vendor site · OpsLevel30 answers · 51 citations · 12 models
vendor site · Guideflow29 answers · 33 citations · 12 models
26 answers · 42 citations · 10 models
vendor site · Cortex26 answers · 36 citations · 12 models
22 answers · 23 citations · 9 models
vendor site · Harness20 answers · 34 citations · 12 models
vendor site · Northflank19 answers · 22 citations · 8 models
19 answers · 19 citations · 10 models
Pages the answers cite
The ten pages named in the most answers, by full address. A page here is one the models returned with a recommendation, not one the index endorses.
33 answers · 12 models
33 answers · 11 models
27 answers · 11 models
27 answers · 10 models
24 answers · 10 models
20 answers · 8 models
19 answers · 10 models
17 answers · 9 models
17 answers · 9 models
14 answers · 6 models
Search against answers
Each company's standing in the answers beside its site's footprint in Google search, one row a site: the products the models named on it with their shares, and the share they add up to; monthly searches on Google, and DataForSEO's estimate of AI search demand (modeled from search signals, directional, not a count of queries to any assistant), for the most-searched of the company's and its products' names (the name is in each row's hover text); estimated monthly organic visits to the site; and its best position in Google's top ten for “best internal developer portal”, “internal developer portal”, “internal developer portals”. US estimates from DataForSEO and Google's Ads Transparency Center. A small company's site, or a mid-sized company's site for its flagship, is marked company; a product on a large parent's site (Google, Microsoft) has no site figures. A column with no figures for this category is left out, and an empty cell means none were seen, not none exist. Two measurements side by side: neither is read as the cause of the other.
| Company and products | Share | Own site cited | Name searches, Google | AI search demand, est. | Organic visits | Paid search a month, est. | Google ads, last 30 days | Best Google position |
|---|---|---|---|---|---|---|---|---|
| Port port.io Port 53%, Port.io 6% | 58% | 166 | 90,500 | 598,288 | 7,729 | 26 | 1 | |
| Backstage backstage.io | 23% | 182 | 110,000 | 7,081 | 2,929 | 2 | ||
| Palo Alto Networks parent site Palo Alto Networks Cortex Cloud 8% | 8% | 227 | 40,500 | 1,088 | ||||
| OpsLevel opslevel.com | 8% | 109 | 320 | 0 | 2,206 | 12 | ||
| Humanitec humanitec.com | 2% | 59 | 260 | 7 | 610 | 0 | ||
| Qovery qovery.com | 2% | 28 | 390 | 4 | 2,459 | 0 | 3 | |
| Roadie roadie.io | 0% | 80 | 110,000 | 6,838 | 416 | 5 | ||
| Configure8 | 0% | 20 | 0 | |||||
| Cycloid cycloid.io | 0% | 24 | 4,400 | 526 | 394 | 13 | ||
| Atlassian atlassian.com Atlassian Compass 0%, Atlassian 0% | 0% | 59 | 720 | 14 | 4,724,388 company | $5,906 | 61 |
Google's top ten for the three searches hold 21 sites; 3 of them are among the sites the models cited here (cortex.io, harness.io, port.io). The first result for “best internal developer portal” is cortex.io.
How they answered
Six framings of the same buying question, each sent to every model in a fresh session with search on. One row per model, so a row shows whether it held its answer under rewording, what it named when cost was the constraint, and what it argued against. Computed from the raw judge labels.ShowHide
How they answered
Six framings of the same buying question, each sent to every model in a fresh session with search on. One row per model, so a row shows whether it held its answer under rewording, what it named when cost was the constraint, and what it argued against. Computed from the raw judge labels.
ShowHideModelDirect“What is the best internal developer portal for a mid-market B2B company?”Paraphrase“Which internal developer platform would you recommend to a mid-sized B2B company?”Comparative“What are the top internal developer portals and how do they differ?”Budget-constrained“What is the best internal developer portal for a company with a limited budget?”Scale-constrained“We are a 500 person company evaluating an internal developer portal. What should we look at?”Negative“Which internal developer portals should I avoid or be cautious about?”
Claude Haiku 4.5no first choiceno first choiceHeldOpsLevel, PortPortno first choicenothing named
Three alternatives
Atlassian Compass, Backstage, Palo Alto Networks Cortex CloudTwo alternatives
Atlassian Compass, BackstageGPT-5.4 miniPalo Alto Networks Cortex CloudHumanitecChangedBackstageagainst: HumanitecBackstageno first choiceagainst: Backstage
Two alternatives
Backstage, PortOne alternative
BackstageFour alternatives
OpsLevel, Palo Alto Networks Cortex Cloud, Port, RoadieFour alternatives
OpsLevel, Palo Alto Networks Cortex Cloud, Port, RoadieGemini 3.5 FlashPortagainst: BackstagePortHeldagainst: Spotify Backstageno first choiceagainst: Atlassian Compass, BackstagePortagainst: Spotify BackstageOpsLevel, Palo Alto Networks Cortex Cloud, Portagainst: Atlassian Compass, Backstageagainst: Atlassian Compass, Datadog Software Catalog, Palo Alto Networks Cortex Cloud, Port, Spotify Backstage
Three alternatives
OpsLevel, Palo Alto Networks Cortex Cloud, RoadieFour alternatives
Northflank, Palo Alto Networks Cortex Cloud, Qovery, RoadieFour alternatives
OpsLevel, Palo Alto Networks Cortex Cloud, Port, RoadieOne alternative
Atlassian CompassOne alternative
RoadiePerplexity SonarPortBackstageChangedBackstageagainst: Mia-PlatformBackstageno first choicenothing named
Two alternatives
Backstage, Palo Alto Networks Cortex CloudThree alternatives
Palo Alto Networks Cortex Cloud, Port, RoadieFive alternatives
Cycloid, Humanitec, OpsLevel, Palo Alto Networks Cortex Cloud, PortThree alternatives
OpsLevel, Port, RoadieGrok 4.1 FastPort.ioPort.ioHeldagainst: Backstage, Humanitec, OpsLevel, RoadieBackstage, PortBackstageOpsLevelagainst: Backstage, OpsLevel, Port
Four alternatives
Atlassian Compass, Backstage, OpsLevel, Palo Alto Networks Cortex CloudOne alternative
Palo Alto Networks Cortex CloudFour alternatives
Harness IDP, OpsLevel, Palo Alto Networks Cortex Cloud, RoadieTwo alternatives
Configure8, Port.ioThree alternatives
Backstage, Palo Alto Networks Cortex Cloud, RoadieMistral SmallPortagainst: BackstageBackstageChangedBackstageBackstageBackstageagainst: Backstage, Port
One alternative
Palo Alto Networks Cortex CloudThree alternatives
Cycloid, Qovery, RoadieFour alternatives
Atlassian Compass, Cycloid, Port, QoveryTwo alternatives
Configure8, Harness IDPFour alternatives
OpsLevel, Palo Alto Networks Cortex Cloud, Port, RoadieDeepSeek V4 FlashPortagainst: BackstagePortHeldBackstage, PortPortagainst: Backstage, Palo Alto Networks Cortex CloudOpsLevel, Palo Alto Networks Cortex Cloud, Portagainst: Backstageagainst: Atlassian Compass, OpsLevel, Palo Alto Networks Cortex Cloud, Port.io, Self-Hosted Backstage
Two alternatives
OpsLevel, Palo Alto Networks Cortex CloudTwo alternatives
Backstage, HumanitecThree alternatives
Humanitec, OpsLevel, Palo Alto Networks Cortex CloudOne alternative
Atlassian CompassOne alternative
Atlassian CompassLlama 4 Maverickno first choiceBackstageChangedno first choiceBackstage, Palo Alto Networks Cortex Cloud, Portno first choicenothing named
Two alternatives
Qovery, RoadieQwen 3.7 FlashPort.ioagainst: BackstagePortChangedagainst: BackstageSpotify's BackstageBackstage, PortPortagainst: Backstageagainst: Confluence, SharePoint, Spotify Backstage
Two alternatives
Configure8, OpsLevelOne alternative
HumanitecThree alternatives
Atlassian, Harness, SpaceliftTwo alternatives
DevPortal, OstiaTwo alternatives
OpsLevel, Palo Alto Networks Cortex CloudKimi K2Portagainst: BackstageOpsLevel, PortChangedagainst: BackstagePortPortno first choiceagainst: Backstage, Humanitec, OpsLevel, Palo Alto Networks Cortex Cloud
Four alternatives
Configure8, OpsLevel, Palo Alto Networks Cortex Cloud, RoadieTwo alternatives
Compass, RoadieFour alternatives
Backstage, OpsLevel, Palo Alto Networks Cortex Cloud, RoadieTwo alternatives
Backstage, RoadieFive alternatives
Backstage, Configure8, OpsLevel, Palo Alto Networks Cortex Cloud, PortGLM 4.7 FlashXPortQoveryChangedagainst: BackstageBackstageBackstageno first choiceagainst: Backstage, OpsLevel, Palo Alto Networks Cortex Cloudagainst: Backstage
Three alternatives
Backstage, OpsLevel, Palo Alto Networks Cortex CloudOne alternative
PortFour alternatives
Configure8, Devtron, OpsLevel, Palo Alto Networks Cortex CloudThree alternatives
ReadMe, Roadie, ZudokuMiniMax M2.5Portagainst: BackstagePortHeldagainst: BackstageBackstagePortno first choiceagainst: Backstageagainst: Backstage
Three alternatives
Configure8, OpsLevel, Palo Alto Networks Cortex CloudOne alternative
Palo Alto Networks Cortex CloudThree alternatives
Atlassian Compass, Configure8, PortTwo alternatives
Backstage, OpsLevelBold is the first choiceAlternatives are counted; the count opens them.What the answer argued against
The record
One row per call: the version string exactly as returned, whether the model searched, sources cited, and latency. Full answer text is in the free responses file. Download the record
Seventy-two rows: every prompt, every model, every answer.
| Prompt | Model | Version string | Time (UTC) | Searched | Sources | Latency |
|---|---|---|---|---|---|---|
| Direct recommendation | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-05 01:50 | no | 0 | 4 s |
| Direct recommendation | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-04 22:50 | yes | 3 | 5 s |
| Direct recommendation | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-04 23:24 | yes | 15 | 20 s |
| Direct recommendation | Perplexity Sonar | sonar | 2026-10-04 23:11 | yes | 16 | 3 s |
| Direct recommendation | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-05 01:09 | yes | 22 | 9 s |
| Direct recommendation | Mistral Small | mistral/mistral-small via mistral | 2026-10-04 23:14 | yes | 10 | 6 s |
| Direct recommendation | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-04 22:18 | yes | 20 | 18 s |
| Direct recommendation | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-05 01:09 | yes | 5 | 2 s |
| Direct recommendation | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-04 22:15 | yes | 10 | 20 s |
| Direct recommendation | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-05 00:28 | yes | 15 | 20 s |
| Direct recommendation | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 00:26 | yes | 25 | 32 s |
| Direct recommendation | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-05 00:34 | yes | 12 | 27 s |
| Paraphrase | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-04 23:09 | no | 0 | 4 s |
| Paraphrase | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-04 23:56 | yes | 2 | 5 s |
| Paraphrase | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-05 01:07 | yes | 7 | 22 s |
| Paraphrase | Perplexity Sonar | sonar | 2026-10-05 00:05 | yes | 16 | 3 s |
| Paraphrase | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-05 00:56 | yes | 24 | 9 s |
| Paraphrase | Mistral Small | mistral/mistral-small via mistral | 2026-10-05 00:23 | yes | 5 | 4 s |
| Paraphrase | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-04 23:14 | yes | 24 | 26 s |
| Paraphrase | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-04 22:57 | yes | 5 | 3 s |
| Paraphrase | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-05 01:12 | no | 0 | 29 s |
| Paraphrase | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-04 22:54 | yes | 18 | 19 s |
| Paraphrase | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 00:16 | yes | 22 | 24 s |
| Paraphrase | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-04 23:25 | yes | 9 | 23 s |
| Comparative | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-05 00:38 | yes | 16 | 9 s |
| Comparative | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-05 01:37 | yes | 5 | 7 s |
| Comparative | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-04 23:08 | yes | 16 | 32 s |
| Comparative | Perplexity Sonar | sonar | 2026-10-05 01:23 | yes | 20 | 5 s |
| Comparative | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-05 00:30 | yes | 16 | 11 s |
| Comparative | Mistral Small | mistral/mistral-small via mistral | 2026-10-05 00:28 | yes | 13 | 9 s |
| Comparative | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-04 22:42 | yes | 25 | 41 s |
| Comparative | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-05 00:01 | yes | 5 | 4 s |
| Comparative | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-04 22:31 | yes | 10 | 18 s |
| Comparative | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-04 22:41 | yes | 14 | 24 s |
| Comparative | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 01:28 | yes | 23 | 53 s |
| Comparative | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-05 00:44 | yes | 10 | 14 s |
| Budget constrained | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-05 00:14 | yes | 14 | 10 s |
| Budget constrained | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-05 01:26 | yes | 5 | 6 s |
| Budget constrained | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-05 00:44 | yes | 16 | 23 s |
| Budget constrained | Perplexity Sonar | sonar | 2026-10-05 01:52 | yes | 16 | 2 s |
| Budget constrained | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-04 23:51 | yes | 20 | 10 s |
| Budget constrained | Mistral Small | mistral/mistral-small via mistral | 2026-10-05 01:08 | yes | 8 | 4 s |
| Budget constrained | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-05 01:52 | yes | 18 | 31 s |
| Budget constrained | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-05 01:16 | yes | 5 | 2 s |
| Budget constrained | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-04 22:28 | yes | 9 | 20 s |
| Budget constrained | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-05 01:35 | yes | 19 | 46 s |
| Budget constrained | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 00:06 | yes | 24 | 43 s |
| Budget constrained | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-04 23:49 | yes | 10 | 18 s |
| Scale constrained | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-05 01:02 | yes | 18 | 12 s |
| Scale constrained | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-05 00:54 | no | 0 | 4 s |
| Scale constrained | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-05 00:55 | yes | 23 | 35 s |
| Scale constrained | Perplexity Sonar | sonar | 2026-10-05 01:37 | yes | 20 | 4 s |
| Scale constrained | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-05 00:57 | yes | 14 | 11 s |
| Scale constrained | Mistral Small | mistral/mistral-small via mistral | 2026-10-05 01:21 | no | 0 | 8 s |
| Scale constrained | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-05 01:25 | yes | 16 | 50 s |
| Scale constrained | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-05 00:55 | yes | 5 | 2 s |
| Scale constrained | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-05 00:42 | yes | 5 | 27 s |
| Scale constrained | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-05 01:10 | yes | 10 | 16 s |
| Scale constrained | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 00:16 | yes | 20 | 43 s |
| Scale constrained | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-05 01:13 | yes | 5 | 11 s |
| Negative framing | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-04 23:17 | no | 0 | 3 s |
| Negative framing | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-04 23:06 | yes | 3 | 4 s |
| Negative framing | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-04 23:50 | yes | 14 | 22 s |
| Negative framing | Perplexity Sonar | sonar | 2026-10-04 23:53 | yes | 20 | 5 s |
| Negative framing | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-04 23:42 | yes | 24 | 9 s |
| Negative framing | Mistral Small | mistral/mistral-small via mistral | 2026-10-05 00:34 | yes | 5 | 6 s |
| Negative framing | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-04 22:30 | yes | 24 | 34 s |
| Negative framing | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-04 22:46 | yes | 5 | 3 s |
| Negative framing | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-05 01:28 | yes | 13 | 34 s |
| Negative framing | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-05 01:52 | yes | 24 | 29 s |
| Negative framing | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-04 22:44 | yes | 22 | 37 s |
| Negative framing | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-04 23:00 | yes | 15 | 28 s |
Normalization in this category
Every judgment call made between the raw labels and the numbers above, listed so it is visible and reversible.
ShowHide
Category-scoped readings
Harness Internal Developer Portal read as Harness IDP
Unresolved, counted raw
Datadog Software Catalog
DevPortal
Devtron
GitHub Projects/Actions
Google Cloud Deploy
Mogenius
Ostia
PlatformOne
Rookout
Upbound
Zudoku
Discontinued, still offered
No shut-down product was recommended here.