Server definition
- Hash
- sha256:e781c70a43375e427210225b1e4689749f70ba393ecefbe0c5e1b00f386bdd00
- What it is
- What a remote MCP server returned when asked what it offers: 5 tools
The blob, as servednamed by its sha256
{
"instructions": "Resolve hardware and model names to ids with search_catalog before calling check_hardware_fit, models_for_hardware, compare_hardware or get_model_specs. Quote the returned url when citing a result. Report a null field as unknown and do not estimate it yourself. Speeds are ranges with a confidence label; memory residency uses all parameters and speed uses active parameters.",
"tools": [
{
"description": "Use this when the user asks whether one specific model runs on one specific machine, at what memory cost, and roughly how fast. Needs a hardware id and a model id from search_catalog. Returns a fit verdict, the memory arithmetic, and a decode-speed range with a confidence label. Memory is sized from all parameters, speed from active parameters. It does not recommend purchases and does not run anything.",
"inputSchema": {
"properties": {
"context_length": {
"default": 4096,
"description": "Context window in tokens that the KV cache is sized for. 512 to 262144. Default 4096.",
"maximum": 262144,
"minimum": 512,
"type": "integer"
},
"hardware_id": {
"description": "Hardware id from search_catalog.",
"maxLength": 80,
"minLength": 1,
"pattern": "^[a-z0-9][a-z0-9._-]*$",
"type": "string"
},
"model_id": {
"description": "Model id from search_catalog.",
"maxLength": 80,
"minLength": 1,
"pattern": "^[a-z0-9][a-z0-9._-]*$",
"type": "string"
},
"quantisation": {
"default": "Q4_K_M",
"description": "Weight quantisation. One of Q2_K, Q3_K_M, Q4_K_M, Q5_K_M, Q6_K, Q8_0, F16. Default Q4_K_M.",
"enum": [
"Q2_K",
"Q3_K_M",
"Q4_K_M",
"Q5_K_M",
"Q6_K",
"Q8_0",
"F16"
],
"type": "string"
}
},
"required": [
"hardware_id",
"model_id"
],
"type": "object"
},
"name": "check_hardware_fit",
"outputSchema": {
"additionalProperties": false,
"properties": {
"context_length": {
"maximum": 9007199254740991,
"minimum": -9007199254740991,
"type": "integer"
},
"data_as_of": {
"description": "Date the bundled catalogue was last updated. Not the time of this call.",
"pattern": "^\\d{4}-\\d{2}-\\d{2}$",
"type": "string"
},
"hardware": {
"additionalProperties": false,
"properties": {
"bandwidth_gb_s": {
"description": "Theoretical peak memory bandwidth.",
"exclusiveMinimum": 0,
"type": "number"
},
"id": {
"maxLength": 80,
"minLength": 1,
"pattern": "^[a-z0-9][a-z0-9._-]*$",
"type": "string"
},
"memory_gb": {
"exclusiveMinimum": 0,
"type": "number"
},
"memory_kind": {
"enum": [
"dedicated",
"unified",
"shared-system"
],
"type": "string"
},
"name": {
"maxLength": 120,
"type": "string"
},
"status": {
"description": "announced means not yet on sale; eol means no longer made.",
"enum": [
"current",
"eol",
"used-market",
"announced"
],
"type": "string"
}
},
"required": [
"id",
"name",
"memory_gb",
"memory_kind",
"bandwidth_gb_s",
"status"
],
"type": "object"
},
"memory": {
"additionalProperties": false,
"description": "Memory arithmetic for the requested quantisation and context length.",
"properties": {
"available_gb": {
"description": "Memory usable for the model after any operating-system reserve.",
"minimum": 0,
"type": "number"
},
"confidence": {
"enum": [
"measured",
"estimated",
"community"
],
"type": "string"
},
"headroom_gb": {
"description": "available - needed. Negative when the model does not fit.",
"type": "number"
},
"kv_basis": {
"description": "Whether the KV cache is computed from a published config or inferred from the parameter count.",
"enum": [
"published_architecture",
"inferred_architecture"
],
"type": "string"
},
"kv_cache_gb": {
"minimum": 0,
"type": "number"
},
"needed_gb": {
"description": "weights + KV cache + overhead.",
"minimum": 0,
"type": "number"
},
"offloaded_gb": {
"description": "Weights that would live in system RAM. 0 when fully resident.",
"minimum": 0,
"type": "number"
},
"overhead_gb": {
"description": "Fixed framework overhead.",
"minimum": 0,
"type": "number"
},
"weights_gb": {
"description": "All parameters quantised. For a Mixture-of-Experts model this counts every expert.",
"minimum": 0,
"type": "number"
}
},
"required": [
"weights_gb",
"kv_cache_gb",
"overhead_gb",
"needed_gb",
"available_gb",
"headroom_gb",
"offloaded_gb",
"confidence",
"kv_basis"
],
"type": "object"
},
"model": {
"additionalProperties": false,
"properties": {
"active_params_b": {
"description": "Parameters read per token in billions. Equals total for dense models. Sets speed, never memory.",
"exclusiveMinimum": 0,
"type": "number"
},
"id": {
"maxLength": 80,
"minLength": 1,
"pattern": "^[a-z0-9][a-z0-9._-]*$",
"type": "string"
},
"is_moe": {
"type": "boolean"
},
"name": {
"maxLength": 120,
"type": "string"
},
"total_params_b": {
"description": "Total parameters in billions. Sets memory residency.",
"exclusiveMinimum": 0,
"type": "number"
}
},
"required": [
"id",
"name",
"total_params_b",
"active_params_b",
"is_moe"
],
"type": "object"
},
"notes": {
"items": {
"maxLength": 300,
"type": "string"
},
"maxItems": 8,
"type": "array"
},
"quantisation": {
"enum": [
"Q2_K",
"Q3_K_M",
"Q4_K_M",
"Q5_K_M",
"Q6_K",
"Q8_0",
"F16"
],
"type": "string"
},
"sources": {
"description": "Primary sources for the model and hardware records. May be empty.",
"items": {
"pattern": "^https:\\/\\/",
"type": "string"
},
"maxItems": 8,
"type": "array"
},
"speed": {
"anyOf": [
{
"additionalProperties": false,
"description": "Decode speed in tokens per second, always a range with a confidence label.",
"properties": {
"basis": {
"anyOf": [
{
"maxLength": 200,
"type": "string"
},
{
"type": "null"
}
],
"description": "Where the figure comes from, e.g. a benchmark label. null for a modelled estimate."
},
"confidence": {
"enum": [
"measured",
"estimated",
"community"
],
"type": "string"
},
"high_tok_s": {
"minimum": 0,
"type": "number"
},
"last_verified": {
"anyOf": [
{
"pattern": "^\\d{4}-\\d{2}-\\d{2}$",
"type": "string"
},
{
"type": "null"
}
]
},
"low_tok_s": {
"minimum": 0,
"type": "number"
},
"sample_count": {
"anyOf": [
{
"exclusiveMinimum": 0,
"maximum": 9007199254740991,
"type": "integer"
},
{
"type": "null"
}
],
"description": "Number of measured runs behind a measured or community figure."
}
},
"required": [
"low_tok_s",
"high_tok_s",
"confidence",
"basis",
"sample_count",
"last_verified"
],
"type": "object"
},
{
"type": "null"
}
],
"description": "null when the model does not fit, because a speed for hardware that cannot load the weights would be invented."
},
"summary": {
"description": "One or two plain-language sentences stating the answer.",
"maxLength": 600,
"type": "string"
},
"url": {
"pattern": "^https:\\/\\/llmconfigurator\\.com\\/en\\/",
"type": "string"
},
"verdict": {
"enum": [
"fits_comfortably",
"fits_tight",
"needs_offload",
"does_not_fit"
],
"type": "string"
}
},
"required": [
"data_as_of",
"summary",
"verdict",
"quantisation",
"context_length",
"model",
"hardware",
"memory",
"speed",
"notes",
"sources",
"url"
],
"type": "object"
}
},
{
"description": "Use this when the user compares two to four machines. Needs hardware ids from search_catalog. Returns memory, bandwidth, memory type, release year, status and a dated price where one is recorded, side by side. A price of null means none is on record. It does not check any model against the machines; use check_hardware_fit for that.",
"inputSchema": {
"properties": {
"hardware_ids": {
"description": "Two to four hardware ids from search_catalog.",
"items": {
"maxLength": 80,
"minLength": 1,
"pattern": "^[a-z0-9][a-z0-9._-]*$",
"type": "string"
},
"maxItems": 4,
"minItems": 2,
"type": "array"
}
},
"required": [
"hardware_ids"
],
"type": "object"
},
"name": "compare_hardware",
"outputSchema": {
"additionalProperties": false,
"properties": {
"data_as_of": {
"description": "Date the bundled catalogue was last updated. Not the time of this call.",
"pattern": "^\\d{4}-\\d{2}-\\d{2}$",
"type": "string"
},
"hardware": {
"items": {
"additionalProperties": false,
"properties": {
"architecture": {
"maxLength": 80,
"type": "string"
},
"bandwidth_gb_s": {
"description": "Theoretical peak memory bandwidth.",
"exclusiveMinimum": 0,
"type": "number"
},
"class": {
"maxLength": 40,
"type": "string"
},
"id": {
"maxLength": 80,
"minLength": 1,
"pattern": "^[a-z0-9][a-z0-9._-]*$",
"type": "string"
},
"memory_gb": {
"exclusiveMinimum": 0,
"type": "number"
},
"memory_kind": {
"enum": [
"dedicated",
"unified",
"shared-system"
],
"type": "string"
},
"name": {
"maxLength": 120,
"type": "string"
},
"notes": {
"anyOf": [
{
"maxLength": 400,
"type": "string"
},
{
"type": "null"
}
]
},
"price": {
"anyOf": [
{
"additionalProperties": false,
"properties": {
"as_of": {
"description": "Date this price was checked.",
"pattern": "^\\d{4}-\\d{2}-\\d{2}$",
"type": "string"
},
"market": {
"enum": [
"new",
"used",
"refurb"
],
"type": "string"
},
"source": {
"maxLength": 200,
"type": "string"
},
"stale": {
"description": "True when the price is older than 90 days.",
"type": "boolean"
},
"usd": {
"exclusiveMinimum": 0,
"type": "number"
}
},
"required": [
"usd",
"as_of",
"market",
"stale",
"source"
],
"type": "object"
},
{
"type": "null"
}
],
"description": "null when this catalogue holds no dated price. Never 0."
},
"release_year": {
"anyOf": [
{
"maximum": 9007199254740991,
"minimum": -9007199254740991,
"type": "integer"
},
{
"type": "null"
}
]
},
"sources": {
"items": {
"pattern": "^https:\\/\\/",
"type": "string"
},
"maxItems": 8,
"type": "array"
},
"status": {
"description": "announced means not yet on sale; eol means no longer made.",
"enum": [
"current",
"eol",
"used-market",
"announced"
],
"type": "string"
},
"system_ram_gb": {
"description": "System RAM available for offload. 0 for unified-memory machines.",
"minimum": 0,
"type": "number"
},
"url": {
"pattern": "^https:\\/\\/llmconfigurator\\.com\\/en\\/",
"type": "string"
}
},
"required": [
"id",
"name",
"memory_gb",
"memory_kind",
"bandwidth_gb_s",
"status",
"class",
"architecture",
"release_year",
"system_ram_gb",
"price",
"notes",
"sources",
"url"
],
"type": "object"
},
"maxItems": 4,
"minItems": 2,
"type": "array"
},
"summary": {
"description": "One or two plain-language sentences stating the answer.",
"maxLength": 600,
"type": "string"
}
},
"required": [
"data_as_of",
"summary",
"hardware"
],
"type": "object"
}
},
{
"description": "Use this when the user asks about one model's size, architecture, context window or licence. Needs a model id from search_catalog. Returns total and active parameters, weight size per quantisation, context length, licence, release date and sources. It does not say whether the model fits any machine; use check_hardware_fit for that.",
"inputSchema": {
"properties": {
"model_id": {
"description": "Model id from search_catalog.",
"maxLength": 80,
"minLength": 1,
"pattern": "^[a-z0-9][a-z0-9._-]*$",
"type": "string"
}
},
"required": [
"model_id"
],
"type": "object"
},
"name": "get_model_specs",
"outputSchema": {
"additionalProperties": false,
"properties": {
"data_as_of": {
"description": "Date the bundled catalogue was last updated. Not the time of this call.",
"pattern": "^\\d{4}-\\d{2}-\\d{2}$",
"type": "string"
},
"model": {
"additionalProperties": false,
"properties": {
"active_params_b": {
"description": "Parameters read per token in billions. Equals total for dense models. Sets speed, never memory.",
"exclusiveMinimum": 0,
"type": "number"
},
"context_length": {
"anyOf": [
{
"exclusiveMinimum": 0,
"maximum": 9007199254740991,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The model's default context window in tokens."
},
"family": {
"maxLength": 120,
"type": "string"
},
"id": {
"maxLength": 80,
"minLength": 1,
"pattern": "^[a-z0-9][a-z0-9._-]*$",
"type": "string"
},
"is_moe": {
"type": "boolean"
},
"license": {
"anyOf": [
{
"maxLength": 80,
"type": "string"
},
{
"type": "null"
}
]
},
"name": {
"maxLength": 120,
"type": "string"
},
"release_date": {
"anyOf": [
{
"pattern": "^\\d{4}-\\d{2}-\\d{2}$",
"type": "string"
},
{
"type": "null"
}
]
},
"status": {
"enum": [
"current",
"legacy",
"deprecated"
],
"type": "string"
},
"total_params_b": {
"description": "Total parameters in billions. Sets memory residency.",
"exclusiveMinimum": 0,
"type": "number"
}
},
"required": [
"id",
"name",
"total_params_b",
"active_params_b",
"is_moe",
"family",
"context_length",
"license",
"release_date",
"status"
],
"type": "object"
},
"quantisations": {
"description": "Weight size per quantisation. Excludes KV cache and overhead, which depend on context length.",
"items": {
"additionalProperties": false,
"properties": {
"bits_per_weight": {
"exclusiveMinimum": 0,
"type": "number"
},
"confidence": {
"enum": [
"measured",
"estimated",
"community"
],
"type": "string"
},
"quantisation": {
"enum": [
"Q2_K",
"Q3_K_M",
"Q4_K_M",
"Q5_K_M",
"Q6_K",
"Q8_0",
"F16"
],
"type": "string"
},
"weights_gb": {
"minimum": 0,
"type": "number"
}
},
"required": [
"quantisation",
"bits_per_weight",
"weights_gb",
"confidence"
],
"type": "object"
},
"type": "array"
},
"sources": {
"items": {
"pattern": "^https:\\/\\/",
"type": "string"
},
"maxItems": 8,
"type": "array"
},
"summary": {
"description": "One or two plain-language sentences stating the answer.",
"maxLength": 600,
"type": "string"
},
"url": {
"pattern": "^https:\\/\\/llmconfigurator\\.com\\/en\\/",
"type": "string"
}
},
"required": [
"data_as_of",
"summary",
"model",
"quantisations",
"sources",
"url"
],
"type": "object"
}
},
{
"description": "Use this when the user asks what models a given machine can run. Needs a hardware id from search_catalog. Returns up to 20 catalogue models ranked for that machine, each with a fit verdict and a speed range, optionally filtered to one use such as coding. It covers only models in the catalogue, at one quantisation and context length per call, and leaves out models that do not fit.",
"inputSchema": {
"properties": {
"context_length": {
"default": 4096,
"description": "Context window in tokens that the KV cache is sized for. 512 to 262144. Default 4096.",
"maximum": 262144,
"minimum": 512,
"type": "integer"
},
"hardware_id": {
"description": "Hardware id from search_catalog.",
"maxLength": 80,
"minLength": 1,
"pattern": "^[a-z0-9][a-z0-9._-]*$",
"type": "string"
},
"limit": {
"default": 10,
"description": "Maximum models, 1 to 20. Default 10.",
"maximum": 20,
"minimum": 1,
"type": "integer"
},
"quantisation": {
"default": "Q4_K_M",
"description": "Weight quantisation. One of Q2_K, Q3_K_M, Q4_K_M, Q5_K_M, Q6_K, Q8_0, F16. Default Q4_K_M.",
"enum": [
"Q2_K",
"Q3_K_M",
"Q4_K_M",
"Q5_K_M",
"Q6_K",
"Q8_0",
"F16"
],
"type": "string"
},
"use_case": {
"description": "Only list models tagged for this use. One of coding, chat, reasoning, agents, vision. Omit for all models.",
"enum": [
"coding",
"chat",
"reasoning",
"agents",
"vision"
],
"type": "string"
}
},
"required": [
"hardware_id"
],
"type": "object"
},
"name": "models_for_hardware",
"outputSchema": {
"additionalProperties": false,
"properties": {
"context_length": {
"maximum": 9007199254740991,
"minimum": -9007199254740991,
"type": "integer"
},
"data_as_of": {
"description": "Date the bundled catalogue was last updated. Not the time of this call.",
"pattern": "^\\d{4}-\\d{2}-\\d{2}$",
"type": "string"
},
"hardware": {
"additionalProperties": false,
"properties": {
"bandwidth_gb_s": {
"description": "Theoretical peak memory bandwidth.",
"exclusiveMinimum": 0,
"type": "number"
},
"id": {
"maxLength": 80,
"minLength": 1,
"pattern": "^[a-z0-9][a-z0-9._-]*$",
"type": "string"
},
"memory_gb": {
"exclusiveMinimum": 0,
"type": "number"
},
"memory_kind": {
"enum": [
"dedicated",
"unified",
"shared-system"
],
"type": "string"
},
"name": {
"maxLength": 120,
"type": "string"
},
"status": {
"description": "announced means not yet on sale; eol means no longer made.",
"enum": [
"current",
"eol",
"used-market",
"announced"
],
"type": "string"
}
},
"required": [
"id",
"name",
"memory_gb",
"memory_kind",
"bandwidth_gb_s",
"status"
],
"type": "object"
},
"models": {
"items": {
"additionalProperties": false,
"properties": {
"model_id": {
"maxLength": 80,
"minLength": 1,
"pattern": "^[a-z0-9][a-z0-9._-]*$",
"type": "string"
},
"name": {
"maxLength": 120,
"type": "string"
},
"needed_gb": {
"minimum": 0,
"type": "number"
},
"rank": {
"exclusiveMinimum": 0,
"maximum": 9007199254740991,
"type": "integer"
},
"speed": {
"anyOf": [
{
"additionalProperties": false,
"description": "Decode speed in tokens per second, always a range with a confidence label.",
"properties": {
"basis": {
"anyOf": [
{
"maxLength": 200,
"type": "string"
},
{
"type": "null"
}
],
"description": "Where the figure comes from, e.g. a benchmark label. null for a modelled estimate."
},
"confidence": {
"enum": [
"measured",
"estimated",
"community"
],
"type": "string"
},
"high_tok_s": {
"minimum": 0,
"type": "number"
},
"last_verified": {
"anyOf": [
{
"pattern": "^\\d{4}-\\d{2}-\\d{2}$",
"type": "string"
},
{
"type": "null"
}
]
},
"low_tok_s": {
"minimum": 0,
"type": "number"
},
"sample_count": {
"anyOf": [
{
"exclusiveMinimum": 0,
"maximum": 9007199254740991,
"type": "integer"
},
{
"type": "null"
}
],
"description": "Number of measured runs behind a measured or community figure."
}
},
"required": [
"low_tok_s",
"high_tok_s",
"confidence",
"basis",
"sample_count",
"last_verified"
],
"type": "object"
},
{
"type": "null"
}
]
},
"url": {
"pattern": "^https:\\/\\/llmconfigurator\\.com\\/en\\/",
"type": "string"
},
"verdict": {
"enum": [
"fits_comfortably",
"fits_tight",
"needs_offload",
"does_not_fit"
],
"type": "string"
}
},
"required": [
"rank",
"model_id",
"name",
"verdict",
"needed_gb",
"speed",
"url"
],
"type": "object"
},
"type": "array"
},
"quantisation": {
"enum": [
"Q2_K",
"Q3_K_M",
"Q4_K_M",
"Q5_K_M",
"Q6_K",
"Q8_0",
"F16"
],
"type": "string"
},
"summary": {
"description": "One or two plain-language sentences stating the answer.",
"maxLength": 600,
"type": "string"
},
"url": {
"pattern": "^https:\\/\\/llmconfigurator\\.com\\/en\\/",
"type": "string"
},
"use_case": {
"anyOf": [
{
"enum": [
"coding",
"chat",
"reasoning",
"agents",
"vision"
],
"type": "string"
},
{
"type": "null"
}
],
"description": "The use-case filter applied, or null."
}
},
"required": [
"data_as_of",
"summary",
"quantisation",
"context_length",
"use_case",
"hardware",
"models",
"url"
],
"type": "object"
}
},
{
"description": "Use this first when the user names a GPU, Mac, mini PC or language model and you need its id for the other tools. Matches names and common aliases in a fixed catalogue of local-LLM hardware and models. It does not return specs, fit or speed, and it does not search the web.",
"inputSchema": {
"properties": {
"kind": {
"description": "Search hardware or models. Required.",
"enum": [
"hardware",
"model"
],
"type": "string"
},
"limit": {
"default": 5,
"description": "Maximum matches, 1 to 10. Default 5.",
"maximum": 10,
"minimum": 1,
"type": "integer"
},
"query": {
"description": "Hardware or model name as the user wrote it, e.g. \"4090\", \"m4 max 64\", \"llama 3.3 70b\".",
"maxLength": 80,
"minLength": 1,
"pattern": "^[^\\u0000-\\u001f\\u007f]*$",
"type": "string"
}
},
"required": [
"query",
"kind"
],
"type": "object"
},
"name": "search_catalog",
"outputSchema": {
"additionalProperties": false,
"properties": {
"data_as_of": {
"description": "Date the bundled catalogue was last updated. Not the time of this call.",
"pattern": "^\\d{4}-\\d{2}-\\d{2}$",
"type": "string"
},
"kind": {
"enum": [
"hardware",
"model"
],
"type": "string"
},
"matches": {
"items": {
"additionalProperties": false,
"properties": {
"detail": {
"description": "A short factual line, e.g. \"24 GB, 1008 GB/s\" or \"8B dense\".",
"maxLength": 160,
"type": "string"
},
"id": {
"description": "Pass this id to the other tools.",
"maxLength": 80,
"minLength": 1,
"pattern": "^[a-z0-9][a-z0-9._-]*$",
"type": "string"
},
"kind": {
"enum": [
"hardware",
"model"
],
"type": "string"
},
"name": {
"maxLength": 120,
"type": "string"
},
"url": {
"pattern": "^https:\\/\\/llmconfigurator\\.com\\/en\\/",
"type": "string"
}
},
"required": [
"id",
"name",
"kind",
"detail",
"url"
],
"type": "object"
},
"type": "array"
},
"summary": {
"description": "One or two plain-language sentences stating the answer.",
"maxLength": 600,
"type": "string"
}
},
"required": [
"data_as_of",
"summary",
"kind",
"matches"
],
"type": "object"
}
}
]
}Verify it yourself
curl -s https://api.teppi.xyz/v1/evidence/sha256:e781c70a43375e427210225b1e4689749f70ba393ecefbe0c5e1b00f386bdd00 | sha256sum