Internal developer portals for small business buyers
Asked as “internal developer portal”, and as “internal developer platform”, on behalf of a small 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
45%
Clear leader
45Port23Backstage11Port.io21others
45% 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 small B2B company. Ordered by share.
ProductFirst-choice shareNegative rateLabelsQuadrant
Show the three 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
06
07Accepted challengerEndorsed leader
0%First-choice share → · lines at 30% share and 25% negative60%
Key
01Port45%
Port45%02Backstage23%
Backstage23%03OpsLevel2%
OpsLevel2%04Roadie2%
Roadie2%05Atlassian Compass0%
Atlassian Compass0%06Palo Alto Networks Cortex Cloud0%
Palo Alto Networks Cortex Cloud0%07Humanitec0%
Humanitec0%What they warned about
Two of twelve models held their first choice under the paraphrase. Claude Haiku 4.5, GPT-5.4 mini, Gemini 3.5 Flash, Perplexity Sonar, Grok 4.1 Fast, Mistral Small, DeepSeek V4 Flash, Qwen 3.7 Flash, GLM 4.7 FlashX and MiniMax M2.5 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.
Backstage57%
38 of 67 labels negative · 14 of 12 models · 17 hard negative
“Backstage (when self-hosted without dedicated platform team)** **Why to avoid:** - Requires 1.5-3 full-time engineers just for ongoing maintenance” GLM 4.7 FlashX, negative prompt
Humanitec61%
11 of 18 labels negative · 8 of 12 models · 5 hard negative
“Orchestrator-heavy platforms such as Humanitec when self-service provisioning is not yet a real pain point... advises skipping these until that need is clear.” Perplexity Sonar, negative prompt
Palo Alto Networks Cortex Cloud36%
13 of 36 labels negative · 8 of 12 models · 1 hard negative
“the scorecard-driven maturity model is overkill” DeepSeek V4 Flash, paraphrase prompt
OpsLevel27%
9 of 33 labels negative · 5 of 12 models · 2 hard negative
“Why to avoid: They are expensive, complex to set up, and heavily sales-driven.” Qwen 3.7 Flash, 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
75 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. 997 links across 162 sites, every framing counted. Ranked by the number of answers carrying the site or page.
48 answers · 48 citations · 11 models
vendor site · Cycloid47 answers · 79 citations · 12 models
vendor site · Port45 answers · 90 citations · 13 models
30 answers · 30 citations · 10 models
vendor site · Guideflow29 answers · 35 citations · 12 models
29 answers · 29 citations · 12 models
vendor site · OpsLevel25 answers · 45 citations · 11 models
24 answers · 25 citations · 11 models
22 answers · 41 citations · 8 models
vendor site · Harness22 answers · 29 citations · 11 models
20 answers · 20 citations · 9 models
vendor site · Northflank17 answers · 21 citations · 7 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.
48 answers · 11 models
30 answers · 10 models
29 answers · 10 models
28 answers · 11 models
23 answers · 10 models
23 answers · 10 models
23 answers · 11 models
21 answers · 8 models
20 answers · 9 models
16 answers · 8 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 45%, Port.io 11% | 57% | 166 | 90,500 | 598,288 | 7,729 | 26 | 1 | |
| Backstage backstage.io | 23% | 182 | 110,000 | 7,081 | 2,929 | 2 | ||
| Qovery qovery.com | 4% | 28 | 390 | 4 | 2,459 | 0 | 3 | |
| OpsLevel opslevel.com | 2% | 109 | 320 | 0 | 2,206 | 12 | ||
| Roadie roadie.io Roadie 2%, Roadie.io 0% | 2% | 80 | 110,000 | 6,838 | 416 | 5 | ||
| Configure8 | 2% | 20 | 0 | |||||
| Confluence atlassian.com Confluence 2%, Atlassian Compass 0% | 2% | 155 | 165,000 | 6,617 | 4,724,388 | $5,906 | 61 | |
| GitLab gitlab.com | 2% | 218 | 90,500 | 4,844 | 603,028 | $884 | 66 | |
| Notion notion.com | 2% | 153 | 673,000 | 15,608 | 1,061,466 | 52 | ||
| DuploCloud | 2% |
Google's top ten for the three searches hold 21 sites; 2 of them are among the sites the models cited here (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 small B2B company?”Paraphrase“Which internal developer platform would you recommend to a small business that sells to other businesses?”Comparative“What are the top internal developer portals for a small team and how do they differ?”Budget-constrained“What is the best internal developer portal for a small company with a tight budget?”Scale-constrained“We are a 30 person company evaluating an internal developer portal. What should we look at?”Negative“Which internal developer portals should a small business avoid or be cautious about?”
Claude Haiku 4.5Portagainst: BackstageGitLabChangedPortConfluence, Notionno first choiceagainst: Backstage
Two alternatives
Atlassian Compass, GitHubOne alternative
GitHubFour alternatives
Atlassian Compass, Backstage, OpsLevel, Palo Alto Networks Cortex CloudTwo alternatives
Backstage, GitLabGPT-5.4 miniRoadieagainst: BackstageBackstageChangedBackstage, RoadieBackstageagainst: Palo Alto Networks Cortex Cloudno first choiceagainst: Backstage
Two alternatives
Palo Alto Networks Cortex Cloud, PortThree alternatives
GitHub Actions, Humanitec, PortOne alternative
PortGemini 3.5 FlashPortagainst: Spotify BackstageDuploCloudChangedagainst: Backstage, Heroku, Railway, RenderPortagainst: Spotify BackstagePortagainst: BackstageOpsLevelagainst: Palo Alto Networks Cortex Cloud, Spotify Backstageagainst: Palo Alto Networks Cortex Cloud, Port, Spotify Backstage
Three alternatives
Atlassian Compass, OpsLevel, Roadie.ioTwo alternatives
Bunnyshell, QoveryThree alternatives
Atlassian Compass, OpsLevel, RoadieThree alternatives
Atlassian Compass, Encore, NorthflankThree alternatives
Atlassian Compass, Port, RoadiePerplexity SonarBackstageRetoolChangedPortagainst: Backstage, CycloidPortagainst: Backstageno first choiceagainst: Backstage, Humanitec
Three alternatives
OpsLevel, Port, RoadieTwo alternatives
Softr, ToolJetTwo alternatives
OpsLevel, RoadieOne alternative
Atlassian CompassGrok 4.1 FastPort.ioagainst: OpsLevelPortChangedagainst: Backstage, HumanitecPortagainst: BackstageBackstageagainst: Palo Alto Networks Cortex Cloud, Port, RoadieBackstageagainst: Backstage, Humanitec, OpsLevel, Palo Alto Networks Cortex Cloud, Roadie
One alternative
BackstageOne alternative
QoveryThree alternatives
Harness IDP, OpsLevel, RoadieTwo alternatives
Portainer IDP, ToriiThree alternatives
Guideflow, OpsLevel, Port.ioMistral SmallBackstageno first choiceChangedPortBackstageBackstageagainst: Backstage, Humanitec
Three alternatives
Backstage, Harness IDP, Palo Alto Networks Cortex CloudTwo alternatives
OpsLevel, PortDeepSeek V4 FlashPortagainst: Backstage, Palo Alto Networks Cortex CloudPort.ioChangedagainst: Backstage, Humanitec, OpsLevel, Palo Alto Networks Cortex CloudPortagainst: BackstagePort.ioagainst: BackstagePortagainst: Backstage, Humanitec, Kratixagainst: Backstage, Humanitec, Kratix, OpsLevel, Palo Alto Networks Cortex Cloud
Two alternatives
Atlassian Compass, OpsLevelOne alternative
QoveryThree alternatives
Atlassian Compass, OpsLevel, Palo Alto Networks Cortex CloudOne alternative
Atlassian CompassTwo alternatives
OpsLevel, Palo Alto Networks Cortex CloudLlama 4 Maverickno first choiceno first choiceHeldno first choiceBackstageno first choiceagainst: Backstage, Palo Alto Networks Cortex Cloud, Port
One alternative
Palo Alto Networks Cortex CloudQwen 3.7 FlashPortagainst: BackstageSpaceliftChangedagainst: Backstage, Red Hat OpenShiftPort.ioagainst: Spotify BackstagePort.ioagainst: Spotify's BackstagePort.ioagainst: Backstageagainst: Aiven, Humanitec, OpsLevel, Spotify's Backstage
One alternative
Cortex.ioTwo alternatives
GitLab, PortOne alternative
CycloidOne alternative
Palo Alto Networks Cortex CloudKimi K2Portagainst: Backstage, OpsLevel, Palo Alto Networks Cortex CloudPortHeldagainst: BackstagePortagainst: Backstage, OpsLevelBackstagePortagainst: Backstageagainst: OpsLevel, Palo Alto Networks Cortex Cloud, Port, Spotify Backstage
One alternative
QoveryOne alternative
RoadieOne alternative
PortTwo alternatives
OpsLevel, Palo Alto Networks Cortex CloudGLM 4.7 FlashXPortagainst: BackstageQoveryChangedPortPortagainst: HumanitecBackstage, Portagainst: Backstage
One alternative
OpsLevelTwo alternatives
Backstage, PortThree alternatives
Backstage, OpsLevel, Palo Alto Networks Cortex CloudTwo alternatives
Backstage, PortainerOne alternative
KratixMiniMax M2.5Backstage, Port.ioagainst: Harness IDPQoveryChangedPortagainst: BackstagePortConfigure8, Portagainst: Atlassian Compassagainst: Backstage
One alternative
PortThree alternatives
Atlassian Compass, Palo Alto Networks Cortex Cloud, RoadieTwo alternatives
Atlassian Compass, BackstageOne alternative
BackstageBold 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 | 2026-10-04 22:15 | yes | 17 | 9 s |
| Direct recommendation | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-05 00:47 | yes | 3 | 4 s |
| Direct recommendation | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-05 01:38 | yes | 19 | 23 s |
| Direct recommendation | Perplexity Sonar | sonar | 2026-10-05 00:36 | yes | 19 | 3 s |
| Direct recommendation | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-04 22:15 | yes | 23 | 9 s |
| Direct recommendation | Mistral Small | mistral/mistral-small via mistral | 2026-10-04 23:30 | yes | 5 | 3 s |
| Direct recommendation | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-05 01:53 | yes | 22 | 27 s |
| Direct recommendation | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-05 01:27 | yes | 6 | 3 s |
| Direct recommendation | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-05 01:55 | yes | 9 | 23 s |
| Direct recommendation | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-04 22:25 | yes | 10 | 44 s |
| Direct recommendation | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-04 23:50 | yes | 24 | 47 s |
| Direct recommendation | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-05 00:20 | yes | 5 | 10 s |
| Paraphrase | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-05 01:44 | no | 0 | 4 s |
| Paraphrase | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-05 00:45 | no | 0 | 2 s |
| Paraphrase | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-04 23:23 | yes | 14 | 24 s |
| Paraphrase | Perplexity Sonar | sonar | 2026-10-04 23:38 | yes | 18 | 3 s |
| Paraphrase | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-05 00:41 | yes | 22 | 10 s |
| Paraphrase | Mistral Small | mistral/mistral-small via mistral | 2026-10-05 00:37 | no | 0 | 2 s |
| Paraphrase | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-04 23:23 | yes | 24 | 34 s |
| Paraphrase | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-05 01:54 | yes | 5 | 3 s |
| Paraphrase | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-05 00:54 | no | 0 | 29 s |
| Paraphrase | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-05 01:41 | yes | 18 | 18 s |
| Paraphrase | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 00:29 | yes | 24 | 207 s |
| Paraphrase | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-05 01:08 | yes | 13 | 45 s |
| Comparative | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-04 22:13 | yes | 9 | 10 s |
| Comparative | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-04 22:58 | yes | 3 | 5 s |
| Comparative | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-05 01:30 | yes | 16 | 35 s |
| Comparative | Perplexity Sonar | sonar | 2026-10-04 23:52 | yes | 19 | 4 s |
| Comparative | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-05 00:46 | yes | 22 | 9 s |
| Comparative | Mistral Small | mistral/mistral-small via mistral | 2026-10-05 00:47 | yes | 18 | 11 s |
| Comparative | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-05 01:57 | yes | 18 | 40 s |
| Comparative | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-05 01:35 | yes | 5 | 1 s |
| Comparative | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-04 23:00 | yes | 10 | 49 s |
| Comparative | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-04 22:25 | yes | 14 | 22 s |
| Comparative | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 01:58 | yes | 18 | 177 s |
| Comparative | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-05 00:53 | yes | 11 | 15 s |
| Budget constrained | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-05 00:47 | no | 0 | 5 s |
| Budget constrained | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-05 00:43 | yes | 4 | 5 s |
| Budget constrained | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-05 01:39 | yes | 17 | 27 s |
| Budget constrained | Perplexity Sonar | sonar | 2026-10-04 23:57 | yes | 18 | 3 s |
| Budget constrained | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-05 00:17 | yes | 24 | 9 s |
| Budget constrained | Mistral Small | mistral/mistral-small via mistral | 2026-10-05 00:38 | yes | 8 | 5 s |
| Budget constrained | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-05 01:54 | yes | 23 | 31 s |
| Budget constrained | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-05 01:34 | yes | 5 | 3 s |
| Budget constrained | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-04 22:30 | yes | 15 | 46 s |
| Budget constrained | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-05 01:00 | yes | 14 | 22 s |
| Budget constrained | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 01:54 | yes | 20 | 40 s |
| Budget constrained | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-05 01:45 | yes | 10 | 16 s |
| Scale constrained | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-05 01:42 | no | 0 | 6 s |
| Scale constrained | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-05 01:34 | no | 0 | 4 s |
| Scale constrained | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-05 00:51 | yes | 9 | 26 s |
| Scale constrained | Perplexity Sonar | sonar | 2026-10-05 00:45 | yes | 20 | 3 s |
| Scale constrained | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-04 22:26 | yes | 14 | 11 s |
| Scale constrained | Mistral Small | mistral/mistral-small via mistral | 2026-10-04 22:27 | no | 0 | 7 s |
| Scale constrained | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-04 22:26 | yes | 17 | 36 s |
| Scale constrained | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-05 01:44 | yes | 5 | 3 s |
| Scale constrained | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-05 01:15 | no | 0 | 25 s |
| Scale constrained | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-05 01:26 | yes | 25 | 24 s |
| Scale constrained | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 01:24 | yes | 19 | 30 s |
| Scale constrained | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-04 23:48 | yes | 5 | 17 s |
| Negative framing | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-05 00:31 | yes | 9 | 8 s |
| Negative framing | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-05 01:02 | yes | 3 | 5 s |
| Negative framing | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-04 23:41 | yes | 19 | 26 s |
| Negative framing | Perplexity Sonar | sonar | 2026-10-05 00:03 | yes | 20 | 4 s |
| Negative framing | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-04 22:11 | yes | 18 | 9 s |
| Negative framing | Mistral Small | mistral/mistral-small via mistral | 2026-10-04 22:40 | yes | 5 | 4 s |
| Negative framing | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-05 01:56 | yes | 22 | 84 s |
| Negative framing | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-05 01:04 | yes | 5 | 2 s |
| Negative framing | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-05 01:35 | yes | 10 | 31 s |
| Negative framing | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-05 00:04 | yes | 18 | 41 s |
| Negative framing | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 00:41 | yes | 24 | 28 s |
| Negative framing | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-05 01:42 | yes | 7 | 26 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
Harness Internal Developer Portal (IDP) read as Harness IDP
Unresolved, counted raw
Atlassian Suite
Backstage, self-hosted
Bunnyshell
Cortex.io
Datadog Internal Developer Portal
Encore
Harness Platform
Softr
ToolJet
Discontinued, still offered
No shut-down product was recommended here.