How should I compare AI API relay prices?

AI API relay pricing should be evaluated alongside model authenticity, protocol compatibility, token billing, interface stability, and provider profiles. Low price should be the last comparison item, not a substitute for detection evidence.

Why shouldn't I choose a relay just by discount?

The same model name may map to different upstream providers, permission tiers, or protocol wrappers. Some low-price providers may have model downgrades, billing discrepancies, or stability issues. Always verify interface behavior first, then compare prices.

AI API Pricing, Models & Routing Evidence

When the same model costs 3-5x different across relays, the cheapest isn't always the best primary route. Check public price bands, protocol evidence, provider profiles, network status, and leaderboard risk before deciding primary, backup, or avoid.

💰 Calculate monthly bill with cost calculator →

📊 Open data 225,118 probe records · 1,958,294 price history rows · 391 site profiles Agent data pilot →

RULE 1

Check the price band

Min ~ max reflects the real market range. Don't look at just one provider's lowest price.

RULE 2

Check evidence coverage

X/Y providers tested by TokenAPI Scan = data credibility. 🟢 ≥60% means "market-verified."

RULE 3

High coverage × Low price

Same coverage → pick lowest price = best value. Different coverage → pick lowest price within ≥60% coverage.

AI API Routing Intelligence: This page isn't just about finding the lowest price — it combines model pricing, evidence coverage, provider profiles, protocol detection, and network status to help you decide whether a model request is suitable as a primary route, backup, observation target, or should be avoided.

Provider profiles Risk leaderboard Upstream status Generate Router Config → How to use this page

🔥 Trending Models · by popularity score

Kimi K3 74 providers · $0.19~$1.00/M ↓99% vs avg Deep review → Gpt 5.5 98 providers · $0.19~$1.12/M ↓99% vs avg Deep review → Claude Opus 5 91 providers · $3.00~$5.00/M ↓83% vs avg Deep review → Gpt 5.6 Sol 94 providers · $0.10~$0.60/M ↓99% vs avg Deep review → Gpt 5.4 96 providers · $0.30~$2.00/M ↓98% vs avg Deep review → Claude Opus 4 8 95 providers · $0.42~$2.12/M ↓97% vs avg Deep review →

🆕 Recently Added · last 7 days

gpt-6.1-terra-pro:batch gpt-6.1-terra-pro gpt-6.1-terra:batch gpt-6.1-terra

Showing 2617 models (47 with public prices, 121 without)

Filter by provider:

Evidence level:

Total: 1231 · 🟢0 verified · 🟡1231 self-reported · coverage 0%

agnes-2.5-flash 🟡 Self-reported

Min input $0.0000/M

Min output $0.0000/M

📊 Price band $0.0000 - $0.0500/M

Evidence: 1 verified · 3 self-reported 25%

View 4 providers' evidence →

muse-spark-1.2-contributor 🟡 Self-reported

Min input $0.0600/M

Min output $0.1200/M

📊 Price band $0.0600 - $0.1000/M

Evidence: 1 verified · 5 self-reported 17%

View 6 providers' evidence →

muse-spark-1.3-contributor 🟡 Self-reported

Min input $0.0600/M

Min output $0.1200/M

📊 Price band $0.0600 - $75.0000/M

Evidence: 1 verified · 7 self-reported 12%

View 8 providers' evidence →

gpt-5.4-mini-openai-compact 🟡 Self-reported

Min input $0.7500/M

Min output $4.5000/M

📊 Price band $0.7500 - $1.7500/M

Evidence: 1 verified · 8 self-reported 11%

View 9 providers' evidence →

muse-spark-1.1 🟡 Self-reported

Min input $0.7500/M

Min output $2.5500/M

📊 Price band $0.7500 - $1.2500/M

Evidence: 1 verified · 8 self-reported 11%

View 9 providers' evidence →

muse-spark-1.3 🟡 Self-reported

Min input $0.7500/M

Min output $2.5500/M

📊 Price band $0.7500 - $1.2500/M

Evidence: 1 verified · 8 self-reported 11%

View 9 providers' evidence →

📦 Expand 1225 more models (less popular / weaker evidence)

muse-spark-1.2 gemini-3-flash-thinking-128 deepseek-v4.1-flash-0910 gpt-5.6-luna-openai-compact gemini-3.7-flash-tiered gemini-3.8-flash-tiered gemini-3.6-flash-low gemini-3.8-flash-low gemini-3.8-flash-medium gemini-3.6-flash-medium gemini-3.7-flash-low gemini-3.7-flash-medium gpt-5.6-luna-2026-07-09 gemini-3.8-flash-high gemini-3.6-flash-tiered gpt-4o-realtime-preview-2025-06-03 gemini-3.5-flash-low gpt-5-minimal gpt-5.3-codex-xhigh gemini-3.7-flash-high gpt-5.1-codex-mini-2025-11-13 gemini-3.6-flash-high gpt-4o-mini-realtime gpt-4o-mini-realtime-preview-2024-12-17 View all 1231 models (search / filter above)

Total: 802 · 🟢0 verified · 🟡802 self-reported · coverage 0%

gpt-5-openai-compact 🟡 Self-reported

Min input $1.2500/M

Min output $6.2500/M

Evidence: 1 verified · 2 self-reported 33%

View 3 providers' evidence →

o4-mini-all 🟡 Self-reported

Min input $1.1000/M

Min output $4.4000/M

📊 Price band $1.1000 - $75.0000/M

Evidence: 1 verified · 3 self-reported 25%

View 4 providers' evidence →

claude-haiku-4-5-thinking 🟡 Self-reported

Min input $1.0000/M

Min output $5.0000/M

Evidence: 1 verified · 4 self-reported 20%

View 5 providers' evidence →

gemini-3.5-flash-medium 🟡 Self-reported

Min input $1.5000/M

Min output $9.0000/M

Evidence: 1 verified · 4 self-reported 20%

View 5 providers' evidence →

gpt-4o-image 🟡 Self-reported

Min input $1.2500/M

Min output $6.2500/M

📊 Price band $1.2500 - $75.0000/M

Evidence: 1 verified · 6 self-reported 14%

View 7 providers' evidence →

o1-mini-all 🟡 Self-reported

Min input $1.1000/M

Min output $4.4000/M

📊 Price band $1.1000 - $75.0000/M

Evidence: 1 verified · 6 self-reported 14%

View 7 providers' evidence →

📦 Expand 796 more models (less popular / weaker evidence)

o3-all gpt-4o-image-vip o3-mini-high-all claude-sonnet-4-5-thinking claude-sonnet-5-5-thinking gpt-daybreak-blue gpt-5.2-codex-xhigh o3-high o3-medium o3-2025-04-16-medium o3-2025-04-16-high o3-low gpt-5.2-xhigh o3-2025-04-16-low claude-opus-5-5-thinking gemini-3.5-flash-extra-low o3-mini-all gemini-3-pro-high gemini-3.5-flash-high gpt-5.6-terra-openai-compact gpt-5.6-terra-2026-07-09 gemini-3-pro-low gpt-5-all gpt-5-chat-2025-10-03 View all 802 models (search / filter above)

Total: 175 · 🟢0 verified · 🟡175 self-reported · coverage 0%

claude-opus-4-5-thinking 🟡 Self-reported

Min input $5.0000/M

Min output $25.0000/M

📊 Price band $5.0000 - $75.0000/M

Evidence: 1 verified · 4 self-reported 20%

View 5 providers' evidence →

o1-all 🟡 Self-reported

Min input $15.0000/M

Min output $60.0000/M

📊 Price band $15.0000 - $75.0000/M

Evidence: 1 verified · 6 self-reported 14%

View 7 providers' evidence →

o1-preview-all 🟡 Self-reported

Min input $15.0000/M

Min output $60.0000/M

📊 Price band $15.0000 - $75.0000/M

Evidence: 1 verified · 6 self-reported 14%

View 7 providers' evidence →

gpt-image-1-all 🟡 Self-reported

Min input $5.0000/M

Min output $40.0000/M

📊 Price band $5.0000 - $75.0000/M

Evidence: 1 verified · 8 self-reported 11%

View 9 providers' evidence →

gpt-5.6-sol-openai-compact 🟡 Self-reported

Min input $5.0000/M

Min output $30.0000/M

📊 Price band $5.0000 - $75.0000/M

Evidence: 1 verified · 20 self-reported 5%

View 21 providers' evidence →

gpt-5.5-instant 🟡 Self-reported

Min input $5.0000/M

Min output $30.0000/M

📊 Price band $5.0000 - $75.0000/M

Evidence: 1 verified · 20 self-reported 5%

View 21 providers' evidence →

📦 Expand 169 more models (less popular / weaker evidence)

gpt-5.5-2026-04-23 gpt-5.5-2026-04-24 o3-deep-research-2025-06-26 gpt-5.5-pro-2026-04-23 o1-preview-2024-09-12 gpt-4o-realtime-preview-2024-12-17 o3-deep-research claude-opus-5-thinking o1 o1-2024-12-17 o3-pro gpt-4o-realtime claude-opus-4-5 claude-fable-5 claude-fable-5-1 gpt-5.5-pro claude-opus-4-20250514 claude-opus-4-1-20250805-thinking gpt-4o-realtime-preview-2024-10-01 gpt-4-turbo-2024-04-09 moonshot-v1-128k gpt-4-1106 gpt-4-0125 grok-beta View all 175 models (search / filter above)

Total: 392 · 🟢0 verified · 🟡392 self-reported · coverage 0%

o3-pro-all 🟡 Self-reported

Min input $20.0000/M

Min output $80.0000/M

📊 Price band $20.0000 - $75.0000/M

Evidence: 1 verified · 6 self-reported 14%

View 7 providers' evidence →

o1-pro-all 🟡 Self-reported

Min input $75.0000/M

Min output $150.0000/M

📊 Price band $75.0000 - $150.0000/M

Evidence: 1 verified · 6 self-reported 14%

View 7 providers' evidence →

o3-pro-2025-06-10 🟡 Self-reported

Min input $20.0000/M

Min output $80.0000/M

📊 Price band $20.0000 - $75.0000/M

Evidence: 1 verified · 52 self-reported 2%

View 53 providers' evidence →

o1-pro-2025-03-19 🟡 Self-reported

Min input $75.0000/M

Min output $300.0000/M

📊 Price band $75.0000 - $150.0000/M

Evidence: 1 verified · 46 self-reported 2%

View 47 providers' evidence →

o1-pro 🟡 Self-reported

Min input $75.0000/M

Min output $300.0000/M

📊 Price band $75.0000 - $150.0000/M

Evidence: 1 verified · 41 self-reported 2%

View 42 providers' evidence →

gpt-4-0613 🟡 Self-reported

Min input $30.0000/M

Min output $60.0000/M

📊 Price band $30.0000 - $75.0000/M

Evidence: 29 self-reported 0%

View 29 providers' evidence →

📦 Expand 386 more models (less popular / weaker evidence)

text-davinci-edit-001 gpt-realtime-mini-2025-12-15 qwq-72b gpt-4.5-preview-2025-02-27 gpt-4.5 gpt-image-1-2025-04-15 tongyi-xiaomi-analysis-flash tongyi-xiaomi-analysis-pro text-davinci-003 gui-plus gpt-image-1-mini-2025-10-06 gpt-realtime-translate gpt-4-all qwen-image-edit-2509 codeqwen1.5-7b qwen-max-1201 qwen3-asr-flash-realtime-2025-10-27 qwen1.5-1.8b qwen2-0.5b qwen-1.8b-longcontext qwen3-s2s-flash-realtime-2025-09-22 qwen-deep-search-planning qwen1.5-0.5b qwen-image-2.0-2026-03-03 View all 392 models (search / filter above)

⚫ 17 models without public pricing (L1 detected but no price sheet found, click to expand)

These models are claimed by at least one relay provider in TokenAPI Scan L1 probing, but no public price sheet was found. This usually means:

The model is newly released and market pricing hasn't stabilized
Relay providers don't list pricing for older/unpopular models (contact sales)
Some models are enterprise-only and don't have public pricing

⚫ batch-get-gpts

⚫ gpt-5.5-cyber

⚫ claude-4-sonnet

⚫ claude-opus-4-1-all

⚫ claude-opus-4-1-thinking

⚫ claude-opus-4-1-thinking-all

⚫ claude-opus-4-all

⚫ claude-opus-4-thinking-all

⚫ claude-sonnet-4-5-all

⚫ claude-sonnet-4-5-thinking-all

⚫ claude-sonnet-4-all

⚫ claude-sonnet-4-thinking-all

⚫ gemini-2.5-flash-all

⚫ gemini-2.5-flash-image-hd

⚫ gemini-2.5-flash-image-preview-hd

⚫ gemini-2.5-flash-image-preview-vip

⚫ gemini-2.5-flash-image-vip

Want to see pricing for these models? Tell us the specific provider → View live-tested relay list

🏷️ Browse prices by vendor

Browse real-time pricing from each relay by model vendor. Each vendor page aggregates all priced models and the cheapest relay.

OpenAI 576 阿里通义千问 369 Google 359 Anthropic 208 xAI 144 DeepSeek 107 字节跳动 100 Meta 98 智谱 81 Mistral 60 Moonshot 45 百度 36 零一万物 17 阶跃星辰 16 讯飞 14 MiniMax 13 Perplexity 8 腾讯 1

AI API providers currently being tracked and tested, organized by category. Cards below list providers with public pricing data.Each provider shows total detected models and priced models. Click a card to view the full detection report.Delisted providers and providers without pricing yet are kept in the collapsed sections below.

No public pricing yet: 288 providers being tracked (not in current price filter)

These providers are still tracked (model lists probed) but no public price sheet has been captured yet. Open a provider profile for its detection report.

1314MCNone 2rkai6.comNone 5202030None 529961None 5cookie.ccNone 724codes.comNone 7470None 88996None 88CodeNone 91CodeNone Ai CodeqaqNone Ai HybgzsNone ai.17nas.comNone ai.cdn.adNone ai.centos.hkNone ai.chaozhi.chatNone ai.flashtopNone ai.kividas.comNone ai.luminai.ccNone ai.ssgoo.netNone ai.xem8k5.topNone ai.xlyun.proNone ai47apiNone AIAPI.worldNone aiapis.ioNone AicanapiNone AicodemirrorNone aicodemirror.comNone Aicoding ShNone AigocodeNone aimuxrNone aircode.xinNone airouter.mxyhi.comNone aisz.momNone aiwanwu.ccNone AnmoryNone anpin.aiNone AnyrouterNone AokAPINone api.aibinghe.xyzNone api.aijws.comNone api.luccc.ukNone api.tmlab.storeNone api.xaix.meNone api2cn.comNone apibNone apiclaude.ccNone apilinkNone ApimartNone apinova.ccNone apiporterNone ApiyiNone AugmuntNone autorouter.ioNone az.zlvipNone b.aiNone B4u QzzNone baobu.xyzNone bbww.topNone beecode.ccNone BeefAPINone bestmax.ccNone bmai.kun8.vipNone bossclaw.orgNone cc-vibe.comNone cc.anyroute.ioNone cca.maya.todayNone ccfly.codesNone ccgoai.comNone ccode.devNone ccrouter.aiNone chaoye.xyzNone ChatAnyWhereNone chiyicn.comNone cialloapiNone Claude APINone Claude APINone Claude Code HubNone Claude OpusNone ClaudeideNone Claude ChinaNone clauqeNone clawopen.topNone cloud-cn.dataeyes.aiNone cnllmNone Co YesNone Code Wenwen AiNone Code X AioNone code-relay.comNone code.deepask.chatNone code.newcli.comNone code.silkcomNone CodeilabNone codesome.aiNone codexone.aieania.techNone console.claudecomNone Cooper APINone CpassNone CubenceNone dasuapiNone dataeyes.aiNone DDSHubNone ddshub.ccNone derouter.aiNone DhcoderNone Dify APINone discordtool.vipNone DmxapiNone DMXAPINone dragoncode.codesNone DreamgerNone duckcode.cnNone DuckcodingNone dzzi.aiNone easy-token.comNone ejkoy.comNone ekan8.comNone ephone.aiNone euzhi.comNone Expansion ChatNone FastGPT APINone feng.cxNone Fk HshwkNone FlapcodeNone Foxcode HshwkNone Foxcode RjjNone free.learn-ai.topNone freemodel.devNone frontier-intelligence.techNone Fuli HxiNone getezclaw.comNone Gpt ZhizengzengNone gpt.geNone gpt.hslife.funNone gptccNone GPTGod196 GptkeyNone GrowUpAPINone GueaiNone gzxsyNone hack3rx.cnNone hangzhale.comNone hao.aiNone heiyucode.comNone HemaAPINone HenapiNone hepingan.topNone HiAPINone hk.ccplus.siteNone hk.hhnto.topNone hohocodeNone HoldaiNone HongshanNone HuanAPINone HuggingFaceNone iacc.ccNone Instcopilot ApiNone IwhalecloudNone JiekouNone jmyr.netNone KelaiAPINone Koozhan APINone langziapiNone Laozhang.aiNone LconaiNone leleweb.cnNone LemongptNone limiao.netNone lingshuai.ccNone lingsuan.topNone linkai.shopNone LinkAPINone linkfreeNone Lino APINone loptu.netNone lucen.ccNone main-new.codesuc.topNone MkeaiNone MnAPINone modcon.topNone mx.free.codesonline.devNone mzlone.topNone N1nNone NbAPINone nbility.devNone nekocode.aiNone New XychataiNone nextoken.onlineNone OaiproNone One APINone Oneapi PaintbotNone openaeon.aiNone OpenAI HKNone OpenclaudeNone opentk.aiNone OpenTokenNone owlai.techNone PackyAPINone PackyAPINone pixelmuse.shopNone platform.ope.aiNone poixe.comNone PoloAPINone portal.gpteamservices.comNone ppword.cnNone PrismAPINone PrivnodeNone proxy2it.comNone ProxyaiNone qiyiguo.ukNone quqiaiNone right.codesNone rkai6.comNone routerlinkNone rpg8.cnNone Runanytime HxiNone RunAPINone sadaiNone SafeAPINone satuo66.onlineNone sbbbbbbbbb.xyzNone schyler.topNone ShitAPINone shiyutv.cnNone SssaicodeNone Store ZzyusNone sub2.congmingai.comNone sub2luchikey.comNone sudocode.usNone sunskii.comNone SuperAPINone switchbase.vipNone tallius.aiNone terminal.pubNone thestocklist.siteNone timebackward.comNone TIMICCNone tokaifyNone Token APINone token.baogutang.topNone tokenflux.devNone tokenmix.aiNone tokenonly.aiNone tokenporterNone toknex.aiNone TongAPINone toolcode.topNone tryallai.comNone Tu ZiNone UiuiapiNone UniapiNone uniprep.worldNone unity2.aiNone us.usclaude.usNone UU APINone VectorEngine AINone vip.crond.devNone wangcai.cloudNone wangxiangai.orgNone wanlaiNone wasdxxNone whlemon.cnNone worldbase.aiNone Wzw De5None x.dogenet.winNone XairouterNone XaixapiNone xb1520.comNone xbhuizNone xcode.bestNone xpluse.plusNone ykh.aiNone YuegleNone yundu.latNone yunserver.cnNone YunwuNone Yunwu ZeaburNone ZenmuxNone ZhangyuAPINone zivv.proNone Ergou APINone Daodun APINone Anjiu APINone Xiaohuli APINone Xiaorui APINone Chunxue APINone Lingmou AINone Shenma Relay APINone

Historical archive: 7 delisted providers (not in current price filter)

These providers have been delisted from /site and the leaderboard. The price page no longer recommends them as current options — links are kept for historical reference and old URL redirects.

AI Proxy Delisted

Historical models: None

AiTuTu Delisted

Historical models: None

API.xyz Delisted

Historical models: None

Flowise Delisted

Historical models: None

GPTLink Delisted

Historical models: None

Lepton AI Delisted

Historical models: None

OK API Delisted

Historical models: None

Bottom line: The price page compares public rate clues but cannot alone determine whether a provider is reliable. A safer order: first check model authenticity and protocol compatibility, then token billing, provider profiles, and network status — compare prices last.

📊 Price comparison across providers

Price gradient (normalized by input price) Verified (≥3 tests in 30 days) Self-reported (official site, no test) Untracked (detected but no public price)

💡 Reading tips: The same model can cost 3-5x more across providers. Verified = TokenAPI Scan tested within 30 days; Self-reported = official site price, untested. Saving money isn't just about the lowest price — check evidence coverage (% = share of tests in the last 30 days).

Aihubmix gpt-3.5-turbo $0.2080

Aihubmix gpt-3.5-turbo-16k $0.4160

Aihubmix gpt-4 $3.4720

Aihubmix gpt-4-turbo-preview $1.3880

Aihubmix claude-3-sonnet $0.4160

Aihubmix claude-3-opus $2.0830

Aihubmix claude-3-haiku $0.0340

Aihubmix mistral-7b-instruct $0.0690

Anthropic (APAC) claude-opus-5-fast $10.0000

Anthropic (APAC) claude-opus-4.8-fast $10.0000

Anthropic (APAC) claude-opus-4.7-fast $30.0000

Anthropic (APAC) claude-opus-4 $15.0000

Anthropic (APAC) claude-3-haiku $0.2500

Anthropic (APAC) claude-sonnet-4 $3.0000

Anthropic (APAC) claude-sonnet-5.5 $2.0000

Anthropic (APAC) claude-sonnet-5.5:batch $1.0000

Anthropic (APAC) claude-opus-5.5 $4.0000

Anthropic (APAC) claude-opus-5.5:batch $2.0000

The snapshot table is server-rendered first; if the chart loads slowly or fails, you can still see price comparisons.

Provider	Model	Input $/1M	Output $/1M
Aihubmix	gpt-3.5-turbo	$0.2080	$0.2080
Aihubmix	gpt-3.5-turbo-16k	$0.4160	$0.4160
Aihubmix	gpt-4	$3.4720	$6.9440
Aihubmix	gpt-4-turbo-preview	$1.3880	$4.1660

📈 Price trend observations Scraped weekly · accumulating over time

Observe price trends per model — input and output separately to avoid one-off quote bias. Price history reveals pricing patterns, promotional returns, and new version releases.For multi-week trends or batch comparison, select a model and price type below.

Model	Min input	Min output	Coverage
gpt-3.5-turbo	$0.2080	$0.2080	Aihubmix
gpt-3.5-turbo-16k	$0.4160	$0.4160	Aihubmix
gpt-4	$3.4720	$6.9440	Aihubmix
gpt-4-turbo-preview	$1.3880	$4.1660	Aihubmix
claude-3-sonnet	$0.4160	$1.6660	Aihubmix
claude-3-opus	$2.0830	$6.2500	Aihubmix
claude-3-haiku	$0.0340	$0.1380	Aihubmix
mistral-7b-instruct	$0.0690	$0.0690	Aihubmix
claude-opus-5-fast	$10.0000	$50.0000	Anthropic（亚太区）
claude-opus-4.8-fast	$10.0000	$50.0000	Anthropic（亚太区）
claude-opus-4.7-fast	$30.0000	$150.0000	Anthropic（亚太区）
claude-opus-4	$15.0000	$75.0000	Anthropic（亚太区）

📅 Static snapshot · scraped weekly · switch model/type above for more

Last updated: 2026-10-29T07:01:08Z

AI API Pricing, Models & Routing Evidence

🔥 Trending Models · by popularity score

🆕 Recently Added · last 7 days

Total: 1231 · 🟢0 verified · 🟡1231 self-reported · coverage 0%

Total: 802 · 🟢0 verified · 🟡802 self-reported · coverage 0%

Total: 175 · 🟢0 verified · 🟡175 self-reported · coverage 0%

Total: 392 · 🟢0 verified · 🟡392 self-reported · coverage 0%

🏷️ Browse prices by vendor

Chinese-market relays (4 providers)

AI platforms & tools (9 providers)

Open-source platforms (5 providers)

Inference acceleration (2 providers)

Candidate promotion (1 providers)

International relays (4 providers)

claude-relay (70 providers)