Sponsored byFusecodeEnterprise coding API for Claude Code, Codex, and model workflows.

LMSpeed

Latency: 6.25 s
Created At: 12/21/2025
Recharge Rate: ¥0.50 per $1 quota

Features

DrawingTaskData ExportCheck-in

Login Methods

GitHubLinuxDOPasskeyGoogle OAuth

Website

API Endpoints

Endpoint 1Historical / Unverified
https://new.xinjianya.top
Endpoint 2Historical / Unverified

Claim this provider

Verify ownership to unlock provider management features:

Edit provider name, content, and links
Get featured with priority traffic and visibility boost
Display a verified badge to build user trust

Hoshimi Miyabi Public API: Permanently free of charge. (Anyone distributing this site for a fee: may your whole family perish). To prevent abuse, we use a check-in system for public welfare access.

Note: Currently, registration only supports Gmail and QQ mail (case-sensitive).

Our existing channels include: Grok, GPT, NVIDIA, DeepSeek, and more.

Health Check

100%Recent availability

History (72 pts)

PastNow

API Benchmarks & Pricing

Compare 63 model rows across audit recency, latest speed tests, throughput, latency, and per-token pricing.

View all 476 models

Model	Input ($/M)	Output ($/M)	Audit	Speed	Latency
李/DeepSeek-V4-pro-reverse	-	-	—	51.0 t/s	3.36 s
deepseek-ai/deepseek-v4-prodefault	Free	Free	—	—	—
deepseek-v4-prodefault	Free	Free	—	25.7 t/s	14.88 s
英伟达/deepseek-ai/deepseek-v4-prodefault	Free	Free	—	—	—
deepseek-ai/deepseek-v4-flashdefault	Free	Free	—	—	—
Deepseek-V4-Flashdefault	Free	Free	—	46.9 t/s	10.09 s
英伟达/deepseek-ai/deepseek-v4-flashdefault	Free	Free	—	—	—
mimo-v2.5-prodefault	$0.100/M	$0.100/M	—	—	—
deepseek-chat	-	-	—	41.9 t/s	1.46 s
z-ai/glm-5.1default	Free	Free	—	—	—
英伟达/z-ai/glm-5.1	-	-	—	21.6 t/s	9.58 s
grok-4.20-0309	-	-	—	111.4 t/s	1.33 s
grok-4.20-beta	-	-	—	57.6 t/s	1.29 s
英伟达/nvidia/nemotron-3-super-120b-a12b	-	-	—	83.4 t/s	5.26 s
李/z-ai/glm5	-	-	—	12.2 t/s	18.82 s
mistralai/mistral-small-4-119b-2603	-	-	—	221.2 t/s	0.55 s
CPA/gpt-5.4-mini	-	-	—	223.3 t/s	3.41 s
claude-3-5-sonnet-20241022	-	-	—	8.0 t/s	2.74 s
claude-sonnet-4-20250514	-	-	—	8.1 t/s	2.70 s
z-ai/glm5	-	-	—	10.7 t/s	46.30 s
google/gemma-3-27b-it	-	-	—	53.7 t/s	0.57 s
meta/llama-4-maverick-17b-128e-instruct	-	-	—	131.8 t/s	0.48 s
qwen/qwen3-coder-480b-a35b-instruct	-	-	—	18.8 t/s	3.69 s
grok-4-thinking	-	-	—	53.7 t/s	0.56 s
grok-4.1-thinking	-	-	—	65.9 t/s	1.78 s
qwen/qwen3.5-397b-a17b	-	-	—	35.4 t/s	2.67 s
英伟达/qwen/qwen3.5-397b-a17b	-	-	—	21.3 t/s	4.35 s
z-ai/glm4.7	-	-	—	46.5 t/s	32.29 s
CPA/gpt-5.4	-	-	—	55.3 t/s	5.58 s
gpt-5.4	-	-	—	47.1 t/s	5.39 s
nvidia/nemotron-3-nano-30b-a3b	-	-	—	246.9 t/s	1.15 s
grok-imagine-1.0-fast	-	-	—	10.0 t/s	4.80 s
grok-4.1-expert	-	-	—	33.1 t/s	1.05 s
grok-4.1-mini	-	-	—	65.3 t/s	4.46 s
grok-4.1-fast	-	-	—	99.4 t/s	1.37 s
meta/llama-3.1-8b-instruct	-	-	—	200.3 t/s	0.42 s
gpt-5.3-codex	-	-	—	32.1 t/s	5.76 s
李/gpt-5.3-codex	-	-	—	52.3 t/s	1.54 s
google/gemma-3-1b-it	-	-	—	213.2 t/s	0.52 s
meta/llama-3.2-3b-instruct	-	-	—	101.4 t/s	0.65 s
google/gemma-2-9b-it	-	-	—	36.1 t/s	0.52 s
meta/llama-3.2-1b-instruct	-	-	—	149.0 t/s	5.26 s
google/gemma-3n-e4b-it	-	-	—	19.4 t/s	0.60 s
google/gemma-7b	-	-	—	45.5 t/s	0.60 s
igenius/italia_10b_instruct_16k	-	-	—	35.7 t/s	0.49 s
google/gemma-2-27b-it	-	-	—	44.6 t/s	0.55 s
minimaxai/minimax-m2.5	-	-	—	60.0 t/s	11.13 s
李/minimaxai/minimax-m2.5	-	-	—	104.5 t/s	0.91 s
GLM-5	-	-	—	23.1 t/s	39.70 s
kimi-k2.5	-	-	—	19.8 t/s	1.88 s
moonshotai/kimi-k2.5	-	-	—	11.3 t/s	62.40 s
李/moonshotai/kimi-k2.5	-	-	—	41.5 t/s	21.86 s
GLM-4.7-Flash	-	-	—	33.9 t/s	1.57 s
GLM-4.7	-	-	—	39.0 t/s	32.32 s
gpt-5.2	-	-	—	52.3 t/s	2.67 s
gpt-5.1-codex-max	-	-	—	72.0 t/s	7.98 s
deepseek-ai/deepseek-v3.2	-	-	—	21.5 t/s	15.34 s
claude-sonnet-4-5-20250929	-	-	—	7.0 t/s	4.09 s
deepseek-ai/deepseek-v3.1-terminus	-	-	—	23.1 t/s	1.01 s
gpt-5	-	-	—	37.8 t/s	15.84 s
openai/gpt-oss-120b	-	-	—	142.7 t/s	1.29 s
openai/gpt-oss-20b	-	-	—	227.9 t/s	1.47 s
英伟达/openai/gpt-oss-120b	-	-	—	144.0 t/s	0.94 s

Showing 63 of 63 model rows

Recent Test Records

Time	Model	Speed	Latency
May 1, 08:43 PM	deepseek-v4-pro	25.69 tok/s	14.88s
May 1, 08:39 PM	deepseek-v4-flash	46.94 tok/s	10.09s
May 1, 03:55 PM	英伟达/z-ai/glm-5.1	21.65 tok/s	9.58s
Apr 26, 08:56 AM	李/DeepSeek-V4-pro-reverse	51.03 tok/s	3.36s
Apr 13, 10:19 AM	grok-4.20-0309	111.36 tok/s	1.33s
Apr 13, 09:22 AM	GLM-5	23.06 tok/s	39.70s
Apr 12, 03:40 PM	openai/gpt-oss-120b	159.61 tok/s	1.24s
Apr 12, 10:51 AM	gpt-5.4	34.99 tok/s	13.15s
Apr 12, 10:50 AM	deepseek-chat	41.90 tok/s	1.46s
Apr 12, 08:01 AM	gpt-5.3-codex	28.52 tok/s	10.77s

Notes

Health checks: Scope: the 72-hour chart and recent availability measure API connectivity only. Each bar summarizes one hour of checks. Targets: LMSpeed tries the configured health check URL and provider status URL first, then API endpoints derived from known API hosts and recent speed-test base URLs. A website host is considered only when it looks like an API endpoint. Probe steps: each candidate goes through DNS lookup, TCP connection, TLS handshake for HTTPS, and an HTTP HEAD request with redirects followed. Probing stops after the first reachable candidate. Reachable criteria: every required network step must succeed. An HTTP response below 500 is treated as reachable, including 401 because it confirms that an authenticated API endpoint responded, except for statuses classified as blocked. Blocked results: HTTP 403, 429, 521, 525, and 530, plus detected WAF or Cloudflare challenges, are shown as blocked and excluded from availability calculations because LMSpeed cannot determine whether the API itself is down. Model availability: when a dedicated test key is configured, LMSpeed sends an authenticated GET request to a derived /models endpoint and compares returned model IDs with this provider's listed models. These per-model results appear in Models & Pricing and are not included in the provider connectivity percentage. Timeouts: TCP connection, TLS handshake, HTTP connectivity, and model requests each use a 20-second timeout. A full run can take longer when several candidates are tried. Frequency: a background worker checks all providers every 5 minutes by default. The 72-hour chart combines those samples into hourly bars, and the schedule may be changed by the service operator. Limit: automated samples are not an SLA and do not guarantee account quota, every model, every region, or successful completion requests. Check the provider's own status page before making operational decisions.
Domain Rating data is sourced from Ahrefs. It is a 0–100 backlink-based domain strength signal and does not measure API speed or reliability.
Announcements and FAQ are read from this provider's NewAPI status snapshot when available. LMSpeed stores the original content and optional English translations from the provider status source, then shows the localized fields on this page.

Similar API Provider Alternatives to Compare

Compare 星见雅 API alternatives against 6 nearby API providers using 945 LMSpeed signals across shared model coverage, pricing, benchmark speed, uptime, and free-model availability.

Provider	Why compare	Models	Free	Avg price	Speed	30d uptime
星见雅 API api-xinjianya-top Xinjianya API is an OpenAI-compatible API relay at api.xinjianya.top. Service availability may be limited.	Current provider baseline	188	211	N/A	67 tok/s	96.9%
天絮 API tianxu-api Tianxu API provides an AI model relay service with multiple access points and stable connectivity.	Faster measured speed Higher 30-day availability Broader model coverage Same provider category	347	1	$173.42/M	93 tok/s	97.7%
llm-whitedream-top RenRen API runs a New API-powered gateway on llm.whitedream.top for aggregated access to multiple AI models.	Faster measured speed Higher 30-day availability Broader model coverage Same provider category	269	0	$15.73/M	271 tok/s	99.5%
ai-071129-xyz MapleLeaf API runs a New API-powered gateway on ai.071129.xyz for aggregated access to multiple AI models.	Higher 30-day availability More free-model options Broader model coverage Same provider category	263	292	N/A	38 tok/s	99.5%
ai-chuyel-top A free, community-supported API service providing access to various AI models for developers and users.	Faster measured speed Broader model coverage Same provider category	237	192	$0.128/M	171 tok/s	88.4%
91vip-futureppo-top 91VIP is a non-profit API service providing access to various AI models including Codex, Claude Code, and Open Code, with specific unlimited-use groups.	Faster measured speed Broader model coverage Same provider category	234	10	$34.12/M	301 tok/s	0.4%
api-42w-shop 42 Gongyi Station is a LinuxDO community-run LLM API relay built by user 4242 for testing welfare-station performance. Supports Gemini, DeepSeek, GPT models, with Gemini rate-limit 10/min. Compatible with immersive-translate and LDC.	Higher 30-day availability More free-model options Same provider category	160	373	N/A	N/A	97.4%

https://api.xinjianya.top

星见雅 API